Sentiment Analysis/Simple Neural Network for Sentiment Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sentiment Analysis\n",
    "In this notebook we will:-\n",
    "- prepare the review text data\n",
    "- use Bag of Words model to prepare train and test \n",
    "- develop and train MLP(multi-layer perceptron) and use it to make prediction\n",
    "\n",
    "### Movie Review Dataset\n",
    " - http://www.cs.cornell.edu/people/pabo/movie-review-data/review_polarity.tar.gz\n",
    " \n",
    " contains 1000 positive and 1000 negative reviews"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T22:27:12.444022Z",
     "start_time": "2020-03-23T22:27:12.436997Z"
    }
   },
   "outputs": [],
   "source": [
    "from nltk.corpus import stopwords\n",
    "import string\n",
    "import re\n",
    "import os\n",
    "from collections import Counter"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load and clean and save tokens\n",
    "\n",
    "### Cleaning Steps:\n",
    "- split by spaces\n",
    "- remove punctuations\n",
    "- remove words that are not alphabetic\n",
    "- remove stopwords and tokens with length 1\n",
    "\n",
    "### saving tokens\n",
    "- load pos and neg sentiments docs and process the docs\n",
    "- make a vocab from all docs and save it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T22:21:30.122696Z",
     "start_time": "2020-03-23T22:21:30.111384Z"
    }
   },
   "outputs": [],
   "source": [
    "def load_doc(file):\n",
    "    with open(file, 'r') as f:\n",
    "        text = f.read()\n",
    "    return text\n",
    "\n",
    "def clean_doc(text):\n",
    "    tokens = text.split()\n",
    "    re_punc = re.compile('[%s]' % re.escape(string.punctuation))\n",
    "    tokens = [re_punc.sub('', w) for w in tokens]\n",
    "    tokens = [word for word in tokens if word.isalpha()]\n",
    "    stops = set(stopwords.words('english'))\n",
    "    tokens = [w for w in tokens if not w in stops and len(w)>1]\n",
    "    return tokens"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T22:23:17.145447Z",
     "start_time": "2020-03-23T22:23:17.114080Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['films',\n",
       " 'adapted',\n",
       " 'comic',\n",
       " 'books',\n",
       " 'plenty',\n",
       " 'success',\n",
       " 'whether',\n",
       " 'theyre',\n",
       " 'superheroes',\n",
       " 'batman',\n",
       " 'superman',\n",
       " 'spawn',\n",
       " 'geared',\n",
       " 'toward',\n",
       " 'kids',\n",
       " 'casper',\n",
       " 'arthouse',\n",
       " 'crowd',\n",
       " 'ghost',\n",
       " 'world',\n",
       " 'theres',\n",
       " 'never',\n",
       " 'really',\n",
       " 'comic',\n",
       " 'book',\n",
       " 'like',\n",
       " 'hell',\n",
       " 'starters',\n",
       " 'created',\n",
       " 'alan',\n",
       " 'moore',\n",
       " 'eddie',\n",
       " 'campbell',\n",
       " 'brought',\n",
       " 'medium',\n",
       " 'whole',\n",
       " 'new',\n",
       " 'level',\n",
       " 'mid',\n",
       " 'series',\n",
       " 'called',\n",
       " 'watchmen',\n",
       " 'say',\n",
       " 'moore',\n",
       " 'campbell',\n",
       " 'thoroughly',\n",
       " 'researched',\n",
       " 'subject',\n",
       " 'jack',\n",
       " 'ripper',\n",
       " 'would',\n",
       " 'like',\n",
       " 'saying',\n",
       " 'michael',\n",
       " 'jackson',\n",
       " 'starting',\n",
       " 'look',\n",
       " 'little',\n",
       " 'odd',\n",
       " 'book',\n",
       " 'graphic',\n",
       " 'novel',\n",
       " 'pages',\n",
       " 'long',\n",
       " 'includes',\n",
       " 'nearly',\n",
       " 'consist',\n",
       " 'nothing',\n",
       " 'footnotes',\n",
       " 'words',\n",
       " 'dont',\n",
       " 'dismiss',\n",
       " 'film',\n",
       " 'source',\n",
       " 'get',\n",
       " 'past',\n",
       " 'whole',\n",
       " 'comic',\n",
       " 'book',\n",
       " 'thing',\n",
       " 'might',\n",
       " 'find',\n",
       " 'another',\n",
       " 'stumbling',\n",
       " 'block',\n",
       " 'hells',\n",
       " 'directors',\n",
       " 'albert',\n",
       " 'allen',\n",
       " 'hughes',\n",
       " 'getting',\n",
       " 'hughes',\n",
       " 'brothers',\n",
       " 'direct',\n",
       " 'seems',\n",
       " 'almost',\n",
       " 'ludicrous',\n",
       " 'casting',\n",
       " 'carrot',\n",
       " 'top',\n",
       " 'well',\n",
       " 'anything',\n",
       " 'riddle',\n",
       " 'better',\n",
       " 'direct',\n",
       " 'film',\n",
       " 'thats',\n",
       " 'set',\n",
       " 'ghetto',\n",
       " 'features',\n",
       " 'really',\n",
       " 'violent',\n",
       " 'street',\n",
       " 'crime',\n",
       " 'mad',\n",
       " 'geniuses',\n",
       " 'behind',\n",
       " 'menace',\n",
       " 'ii',\n",
       " 'society',\n",
       " 'ghetto',\n",
       " 'question',\n",
       " 'course',\n",
       " 'whitechapel',\n",
       " 'londons',\n",
       " 'east',\n",
       " 'end',\n",
       " 'filthy',\n",
       " 'sooty',\n",
       " 'place',\n",
       " 'whores',\n",
       " 'called',\n",
       " 'unfortunates',\n",
       " 'starting',\n",
       " 'get',\n",
       " 'little',\n",
       " 'nervous',\n",
       " 'mysterious',\n",
       " 'psychopath',\n",
       " 'carving',\n",
       " 'profession',\n",
       " 'surgical',\n",
       " 'precision',\n",
       " 'first',\n",
       " 'stiff',\n",
       " 'turns',\n",
       " 'copper',\n",
       " 'peter',\n",
       " 'godley',\n",
       " 'robbie',\n",
       " 'coltrane',\n",
       " 'world',\n",
       " 'enough',\n",
       " 'calls',\n",
       " 'inspector',\n",
       " 'frederick',\n",
       " 'abberline',\n",
       " 'johnny',\n",
       " 'depp',\n",
       " 'blow',\n",
       " 'crack',\n",
       " 'case',\n",
       " 'abberline',\n",
       " 'widower',\n",
       " 'prophetic',\n",
       " 'dreams',\n",
       " 'unsuccessfully',\n",
       " 'tries',\n",
       " 'quell',\n",
       " 'copious',\n",
       " 'amounts',\n",
       " 'absinthe',\n",
       " 'opium',\n",
       " 'upon',\n",
       " 'arriving',\n",
       " 'whitechapel',\n",
       " 'befriends',\n",
       " 'unfortunate',\n",
       " 'named',\n",
       " 'mary',\n",
       " 'kelly',\n",
       " 'heather',\n",
       " 'graham',\n",
       " 'say',\n",
       " 'isnt',\n",
       " 'proceeds',\n",
       " 'investigate',\n",
       " 'horribly',\n",
       " 'gruesome',\n",
       " 'crimes',\n",
       " 'even',\n",
       " 'police',\n",
       " 'surgeon',\n",
       " 'cant',\n",
       " 'stomach',\n",
       " 'dont',\n",
       " 'think',\n",
       " 'anyone',\n",
       " 'needs',\n",
       " 'briefed',\n",
       " 'jack',\n",
       " 'ripper',\n",
       " 'wont',\n",
       " 'go',\n",
       " 'particulars',\n",
       " 'say',\n",
       " 'moore',\n",
       " 'campbell',\n",
       " 'unique',\n",
       " 'interesting',\n",
       " 'theory',\n",
       " 'identity',\n",
       " 'killer',\n",
       " 'reasons',\n",
       " 'chooses',\n",
       " 'slay',\n",
       " 'comic',\n",
       " 'dont',\n",
       " 'bother',\n",
       " 'cloaking',\n",
       " 'identity',\n",
       " 'ripper',\n",
       " 'screenwriters',\n",
       " 'terry',\n",
       " 'hayes',\n",
       " 'vertical',\n",
       " 'limit',\n",
       " 'rafael',\n",
       " 'yglesias',\n",
       " 'les',\n",
       " 'mis',\n",
       " 'rables',\n",
       " 'good',\n",
       " 'job',\n",
       " 'keeping',\n",
       " 'hidden',\n",
       " 'viewers',\n",
       " 'end',\n",
       " 'funny',\n",
       " 'watch',\n",
       " 'locals',\n",
       " 'blindly',\n",
       " 'point',\n",
       " 'finger',\n",
       " 'blame',\n",
       " 'jews',\n",
       " 'indians',\n",
       " 'englishman',\n",
       " 'could',\n",
       " 'never',\n",
       " 'capable',\n",
       " 'committing',\n",
       " 'ghastly',\n",
       " 'acts',\n",
       " 'hells',\n",
       " 'ending',\n",
       " 'whistling',\n",
       " 'stonecutters',\n",
       " 'song',\n",
       " 'simpsons',\n",
       " 'days',\n",
       " 'holds',\n",
       " 'back',\n",
       " 'electric',\n",
       " 'carwho',\n",
       " 'made',\n",
       " 'steve',\n",
       " 'guttenberg',\n",
       " 'star',\n",
       " 'dont',\n",
       " 'worry',\n",
       " 'itll',\n",
       " 'make',\n",
       " 'sense',\n",
       " 'see',\n",
       " 'onto',\n",
       " 'hells',\n",
       " 'appearance',\n",
       " 'certainly',\n",
       " 'dark',\n",
       " 'bleak',\n",
       " 'enough',\n",
       " 'surprising',\n",
       " 'see',\n",
       " 'much',\n",
       " 'looks',\n",
       " 'like',\n",
       " 'tim',\n",
       " 'burton',\n",
       " 'film',\n",
       " 'planet',\n",
       " 'apes',\n",
       " 'times',\n",
       " 'seems',\n",
       " 'like',\n",
       " 'sleepy',\n",
       " 'hollow',\n",
       " 'print',\n",
       " 'saw',\n",
       " 'wasnt',\n",
       " 'completely',\n",
       " 'finished',\n",
       " 'color',\n",
       " 'music',\n",
       " 'finalized',\n",
       " 'comments',\n",
       " 'marilyn',\n",
       " 'manson',\n",
       " 'cinematographer',\n",
       " 'peter',\n",
       " 'deming',\n",
       " 'dont',\n",
       " 'say',\n",
       " 'word',\n",
       " 'ably',\n",
       " 'captures',\n",
       " 'dreariness',\n",
       " 'victorianera',\n",
       " 'london',\n",
       " 'helped',\n",
       " 'make',\n",
       " 'flashy',\n",
       " 'killing',\n",
       " 'scenes',\n",
       " 'remind',\n",
       " 'crazy',\n",
       " 'flashbacks',\n",
       " 'twin',\n",
       " 'peaks',\n",
       " 'even',\n",
       " 'though',\n",
       " 'violence',\n",
       " 'film',\n",
       " 'pales',\n",
       " 'comparison',\n",
       " 'blackandwhite',\n",
       " 'comic',\n",
       " 'oscar',\n",
       " 'winner',\n",
       " 'martin',\n",
       " 'childs',\n",
       " 'shakespeare',\n",
       " 'love',\n",
       " 'production',\n",
       " 'design',\n",
       " 'turns',\n",
       " 'original',\n",
       " 'prague',\n",
       " 'surroundings',\n",
       " 'one',\n",
       " 'creepy',\n",
       " 'place',\n",
       " 'even',\n",
       " 'acting',\n",
       " 'hell',\n",
       " 'solid',\n",
       " 'dreamy',\n",
       " 'depp',\n",
       " 'turning',\n",
       " 'typically',\n",
       " 'strong',\n",
       " 'performance',\n",
       " 'deftly',\n",
       " 'handling',\n",
       " 'british',\n",
       " 'accent',\n",
       " 'ians',\n",
       " 'holm',\n",
       " 'joe',\n",
       " 'goulds',\n",
       " 'secret',\n",
       " 'richardson',\n",
       " 'dalmatians',\n",
       " 'log',\n",
       " 'great',\n",
       " 'supporting',\n",
       " 'roles',\n",
       " 'big',\n",
       " 'surprise',\n",
       " 'graham',\n",
       " 'cringed',\n",
       " 'first',\n",
       " 'time',\n",
       " 'opened',\n",
       " 'mouth',\n",
       " 'imagining',\n",
       " 'attempt',\n",
       " 'irish',\n",
       " 'accent',\n",
       " 'actually',\n",
       " 'wasnt',\n",
       " 'half',\n",
       " 'bad',\n",
       " 'film',\n",
       " 'however',\n",
       " 'good',\n",
       " 'strong',\n",
       " 'violencegore',\n",
       " 'sexuality',\n",
       " 'language',\n",
       " 'drug',\n",
       " 'content']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# test load_doc and clean_doc\n",
    "filename = 'review_polarity/txt_sentoken/pos/cv000_29590.txt'\n",
    "text = load_doc(filename)\n",
    "tokens = clean_doc(text)\n",
    "tokens"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T22:26:32.712879Z",
     "start_time": "2020-03-23T22:26:32.703668Z"
    }
   },
   "outputs": [],
   "source": [
    "def add_doc_to_vocab(file, vocab):\n",
    "    text = load_doc(file)\n",
    "    tokens = clean_doc(text)\n",
    "    vocab.update(tokens)\n",
    "    \n",
    "def process_docs(folder, vocab):\n",
    "    for filename in os.listdir(folder):\n",
    "        if filename.startswith('cv9'):\n",
    "            continue\n",
    "        path = folder + \"/\" + filename\n",
    "        add_doc_to_vocab(path, vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T22:28:36.479344Z",
     "start_time": "2020-03-23T22:28:31.116801Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Length of vocab is : 44276\n",
      "Most common 50 tokens are:  [('film', 7983), ('one', 4946), ('movie', 4826), ('like', 3201), ('even', 2262), ('good', 2080), ('time', 2041), ('story', 1907), ('films', 1873), ('would', 1844), ('much', 1824), ('also', 1757), ('characters', 1735), ('get', 1724), ('character', 1703), ('two', 1643), ('first', 1588), ('see', 1557), ('way', 1515), ('well', 1511), ('make', 1418), ('really', 1407), ('little', 1351), ('life', 1334), ('plot', 1288), ('people', 1269), ('could', 1248), ('bad', 1248), ('scene', 1241), ('movies', 1238), ('never', 1201), ('best', 1179), ('new', 1140), ('scenes', 1135), ('man', 1131), ('many', 1130), ('doesnt', 1118), ('know', 1092), ('dont', 1086), ('hes', 1024), ('great', 1014), ('another', 992), ('action', 985), ('love', 977), ('us', 967), ('go', 952), ('director', 948), ('end', 946), ('something', 945), ('still', 936)]\n"
     ]
    }
   ],
   "source": [
    "# test add_doc_to_vocab and process_docs\n",
    "vocab = Counter()\n",
    "process_docs(\"review_polarity/txt_sentoken/pos\", vocab)\n",
    "process_docs(\"review_polarity/txt_sentoken/neg\", vocab)\n",
    "print(\"Length of vocab is :\",len(vocab))\n",
    "print(\"Most common 50 tokens are: \",vocab.most_common(50))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T22:30:40.930035Z",
     "start_time": "2020-03-23T22:30:40.914886Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Length of vocab after removing low occuring tokens is : 25767\n"
     ]
    }
   ],
   "source": [
    "# remove low occuring tokens\n",
    "min_occur = 2\n",
    "tokens = [k for k,c in vocab.items() if c>= min_occur]\n",
    "print(\"Length of vocab after removing low occuring tokens is :\",len(tokens))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T22:33:40.265596Z",
     "start_time": "2020-03-23T22:33:40.259597Z"
    }
   },
   "outputs": [],
   "source": [
    "# save list to file \n",
    "def save_list(lines, filename):\n",
    "    with open(filename, 'w') as f:\n",
    "        f.write(\"\\n\".join(lines))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T22:33:41.423877Z",
     "start_time": "2020-03-23T22:33:41.414870Z"
    }
   },
   "outputs": [],
   "source": [
    "save_list(tokens, 'vocab.txt')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bag-of-Words Representation\n",
    "- BOW is a way of extracting features from text so that text input can be used in ML algorithms.\n",
    "\n",
    "- Each doc of positive and negative sentiment will be converted into a vectore repr.\n",
    "\n",
    "- Words in a doc are scored and scores are placed in corresponding location of repr."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:05:37.178998Z",
     "start_time": "2020-03-23T23:05:37.170032Z"
    }
   },
   "outputs": [],
   "source": [
    "def doc_to_line(filename, vocab):\n",
    "    text = load_doc(filename)\n",
    "    tokens = clean_doc(text)\n",
    "    tokens = [w for w in tokens if w in vocab] #remove token which is not in vocab\n",
    "    return ' '.join(tokens)\n",
    "\n",
    "def process_docs(folder, vocab, is_train):\n",
    "    lines = []\n",
    "    for file in os.listdir(folder):\n",
    "        if is_train and file.startswith('cv9'):\n",
    "            continue\n",
    "        if not is_train and not file.startswith('cv9'):\n",
    "            continue\n",
    "        path = folder + \"/\" + file\n",
    "        line = doc_to_line(path, vocab)\n",
    "        lines.append(line)\n",
    "    return lines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:05:37.328861Z",
     "start_time": "2020-03-23T23:05:37.321021Z"
    }
   },
   "outputs": [],
   "source": [
    "# load and clean dataset\n",
    "def load_clean_dataset(vocab, is_train):\n",
    "    pos = process_docs(\"review_polarity/txt_sentoken/pos\", vocab, is_train)\n",
    "    neg = process_docs(\"review_polarity/txt_sentoken/neg\", vocab, is_train)\n",
    "    docs = pos+neg\n",
    "    labels = [1 for _ in range(len(pos))] + [0 for _ in range(len(neg))]\n",
    "    return docs, labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reviews to BOW vectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:05:38.206025Z",
     "start_time": "2020-03-23T23:05:38.200090Z"
    }
   },
   "outputs": [],
   "source": [
    "from keras.preprocessing.text import Tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:05:38.545665Z",
     "start_time": "2020-03-23T23:05:38.537352Z"
    }
   },
   "outputs": [],
   "source": [
    "def create_tokenizer(reviews):\n",
    "    tokenizer  = Tokenizer()\n",
    "    tokenizer.fit_on_texts(reviews)\n",
    "    return tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:15:05.365170Z",
     "start_time": "2020-03-23T23:14:59.585817Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We have 1800 train reviews.\n",
      "We have 200 test reviews.\n"
     ]
    }
   ],
   "source": [
    "#load saved vocab\n",
    "vocab_filename = 'vocab.txt'\n",
    "vocab = load_doc(vocab_filename)\n",
    "vocab = set(vocab.split())\n",
    "\n",
    "train_reviews, y_train = load_clean_dataset(vocab, True)\n",
    "test_reviews, y_test = load_clean_dataset(vocab, False)\n",
    "\n",
    "print(\"We have\", len(train_reviews), \"train reviews.\")\n",
    "print(\"We have\", len(test_reviews), \"test reviews.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:18:23.776237Z",
     "start_time": "2020-03-23T23:18:23.740049Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>entertainment review website coming soon hunte...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>one responses enjoy detroit rock city probably...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>im scientist thats dr alexander mccabe bob gun...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>would one could see well youre super smart bio...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ring probably one creepiest movies ive seen ye...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>movies based video games street fighter mario ...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>understood passion understood calm bud white w...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>almost full decade steven spielbergs saving pr...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>remake la cage aux folles features gay couple ...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>robin williams time without beard returns dram...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   x  y\n",
       "0  entertainment review website coming soon hunte...  0\n",
       "1  one responses enjoy detroit rock city probably...  0\n",
       "2  im scientist thats dr alexander mccabe bob gun...  0\n",
       "3  would one could see well youre super smart bio...  0\n",
       "4  ring probably one creepiest movies ive seen ye...  1\n",
       "5  movies based video games street fighter mario ...  0\n",
       "6  understood passion understood calm bud white w...  1\n",
       "7  almost full decade steven spielbergs saving pr...  1\n",
       "8  remake la cage aux folles features gay couple ...  1\n",
       "9  robin williams time without beard returns dram...  0"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#shuffle data\n",
    "import pandas as pd\n",
    "df = pd.DataFrame({\"x\":train_reviews, \"y\":y_train})\n",
    "df = df.sample(frac=1).reset_index(drop=True)\n",
    "df.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:18:52.066028Z",
     "start_time": "2020-03-23T23:18:52.060930Z"
    }
   },
   "outputs": [],
   "source": [
    "train_reviews, y_train = df[\"x\"], df[\"y\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:18:55.383774Z",
     "start_time": "2020-03-23T23:18:53.640904Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of training data (1800, 25768)\n",
      "Shape of testing data (200, 25768)\n"
     ]
    }
   ],
   "source": [
    "tokenizer = create_tokenizer(train_reviews)\n",
    "\n",
    "#encode data \n",
    "xtrain = tokenizer.texts_to_matrix(train_reviews, mode='freq')\n",
    "xtest = tokenizer.texts_to_matrix(test_reviews, mode='freq')\n",
    "\n",
    "print(\"Shape of training data\", xtrain.shape)\n",
    "print(\"Shape of testing data\", xtest.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating Model\n",
    "- we will make simple MLP model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:09:32.723069Z",
     "start_time": "2020-03-23T23:09:32.715873Z"
    }
   },
   "outputs": [],
   "source": [
    "from keras.layers import Dense\n",
    "from keras.models import Sequential\n",
    "\n",
    "n_words = xtrain.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:13:55.846809Z",
     "start_time": "2020-03-23T23:13:55.837555Z"
    }
   },
   "outputs": [],
   "source": [
    "def Model(n_words):\n",
    "    model = Sequential()\n",
    "    model.add(Dense(50, input_shape=(n_words,), activation=\"relu\"))\n",
    "    model.add(Dense(25, activation=\"relu\"))\n",
    "    model.add(Dense(1, activation=\"sigmoid\"))\n",
    "    model.compile(loss=\"binary_crossentropy\", optimizer=\"adam\", metrics=['accuracy'])\n",
    "    model.summary()\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:22:39.887300Z",
     "start_time": "2020-03-23T23:22:39.808175Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_5\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_13 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_14 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_15 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Model(n_words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:23:23.927786Z",
     "start_time": "2020-03-23T23:22:39.939114Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.6923 - accuracy: 0.5346 - val_loss: 0.6871 - val_accuracy: 0.5444\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.6618 - accuracy: 0.6019 - val_loss: 0.6349 - val_accuracy: 0.8000\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.5128 - accuracy: 0.8568 - val_loss: 0.4752 - val_accuracy: 0.8389\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.2573 - accuracy: 0.9741 - val_loss: 0.3530 - val_accuracy: 0.8556\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.1042 - accuracy: 0.9926 - val_loss: 0.3129 - val_accuracy: 0.8444\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0456 - accuracy: 0.9988 - val_loss: 0.2773 - val_accuracy: 0.8556\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0223 - accuracy: 1.0000 - val_loss: 0.2720 - val_accuracy: 0.8667\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0129 - accuracy: 1.0000 - val_loss: 0.2731 - val_accuracy: 0.8778\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0081 - accuracy: 1.0000 - val_loss: 0.2735 - val_accuracy: 0.8722\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0055 - accuracy: 1.0000 - val_loss: 0.2765 - val_accuracy: 0.8722\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x1a1e07df98>"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(xtrain, y_train, epochs=10, batch_size = 8, validation_split=0.1, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-23T23:23:26.436467Z",
     "start_time": "2020-03-23T23:23:26.357625Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200/200 [==============================] - 0s 344us/step\n",
      "Accuracy on test data is 91.00000262260437\n"
     ]
    }
   ],
   "source": [
    "loss, acc = model.evaluate(xtest, y_test)\n",
    "print(\"Accuracy on test data is\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparing word scoring methods\n",
    "- **binary** :- where words are present mark as 1 else 0\n",
    "- **count** :- occurance count for each word is marked\n",
    "- **tfidf** :- word is scored based on their frequency, words that are common across document are penalized\n",
    "- **freq** :- words are scored based on their frequency of occurance within doc.\n",
    "\n",
    "\n",
    "We will evaluate each mode 10times and accumulate the accuracy each time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-24T00:12:00.890053Z",
     "start_time": "2020-03-23T23:41:05.978421Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_16\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_46 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_47 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_48 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4494 - accuracy: 0.7889 - val_loss: 0.3121 - val_accuracy: 0.8778\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0228 - accuracy: 0.9957 - val_loss: 0.3429 - val_accuracy: 0.8944\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0021 - accuracy: 1.0000 - val_loss: 0.3816 - val_accuracy: 0.8889\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 6.4452e-04 - accuracy: 1.0000 - val_loss: 0.4204 - val_accuracy: 0.8944\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.0221e-04 - accuracy: 1.0000 - val_loss: 0.4469 - val_accuracy: 0.8944\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.6987e-04 - accuracy: 1.0000 - val_loss: 0.4526 - val_accuracy: 0.8889\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.0332e-04 - accuracy: 1.0000 - val_loss: 0.4703 - val_accuracy: 0.8833\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.6451e-05 - accuracy: 1.0000 - val_loss: 0.4846 - val_accuracy: 0.8889\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.4502e-05 - accuracy: 1.0000 - val_loss: 0.4934 - val_accuracy: 0.8889\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 3.0440e-05 - accuracy: 1.0000 - val_loss: 0.5041 - val_accuracy: 0.8889\n",
      "200/200 [==============================] - 0s 461us/step\n",
      "Accuracy on test data for mode binary is 91.50000214576721\n",
      "Model: \"sequential_17\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_49 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_50 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_51 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4730 - accuracy: 0.7735 - val_loss: 0.3083 - val_accuracy: 0.8778\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0404 - accuracy: 0.9932 - val_loss: 0.3404 - val_accuracy: 0.9000\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0040 - accuracy: 1.0000 - val_loss: 0.3511 - val_accuracy: 0.8722\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 9.6619e-04 - accuracy: 1.0000 - val_loss: 0.3733 - val_accuracy: 0.8722\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 3.7161e-04 - accuracy: 1.0000 - val_loss: 0.4018 - val_accuracy: 0.8722\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 2.0821e-04 - accuracy: 1.0000 - val_loss: 0.4098 - val_accuracy: 0.8778\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.3287e-04 - accuracy: 1.0000 - val_loss: 0.4201 - val_accuracy: 0.8722\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 9.1535e-05 - accuracy: 1.0000 - val_loss: 0.4344 - val_accuracy: 0.8722\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.5286e-05 - accuracy: 1.0000 - val_loss: 0.4399 - val_accuracy: 0.8722\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.8106e-05 - accuracy: 1.0000 - val_loss: 0.4461 - val_accuracy: 0.8722\n",
      "200/200 [==============================] - 0s 295us/step\n",
      "Accuracy on test data for mode binary is 90.49999713897705\n",
      "Model: \"sequential_18\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_52 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_53 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_54 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4651 - accuracy: 0.7809 - val_loss: 0.3072 - val_accuracy: 0.8611\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0498 - accuracy: 0.9877 - val_loss: 0.3138 - val_accuracy: 0.8889\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0034 - accuracy: 1.0000 - val_loss: 0.3287 - val_accuracy: 0.8722\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 7.0445e-04 - accuracy: 1.0000 - val_loss: 0.3608 - val_accuracy: 0.8722\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.3665e-04 - accuracy: 1.0000 - val_loss: 0.3870 - val_accuracy: 0.8778\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.6846e-04 - accuracy: 1.0000 - val_loss: 0.4079 - val_accuracy: 0.8833\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 8.9989e-05 - accuracy: 1.0000 - val_loss: 0.4307 - val_accuracy: 0.8833\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 5.3262e-05 - accuracy: 1.0000 - val_loss: 0.4433 - val_accuracy: 0.8722\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 3.5457e-05 - accuracy: 1.0000 - val_loss: 0.4575 - val_accuracy: 0.8778\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 2.3811e-05 - accuracy: 1.0000 - val_loss: 0.4684 - val_accuracy: 0.8778\n",
      "200/200 [==============================] - 0s 339us/step\n",
      "Accuracy on test data for mode binary is 89.49999809265137\n",
      "Model: \"sequential_19\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_55 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_56 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_57 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4573 - accuracy: 0.7870 - val_loss: 0.3135 - val_accuracy: 0.8667\n",
      "Epoch 2/10\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0310 - accuracy: 0.9951 - val_loss: 0.3209 - val_accuracy: 0.8722\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0040 - accuracy: 1.0000 - val_loss: 0.3298 - val_accuracy: 0.8778\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0013 - accuracy: 1.0000 - val_loss: 0.3459 - val_accuracy: 0.8833\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.3941e-04 - accuracy: 1.0000 - val_loss: 0.3590 - val_accuracy: 0.8833\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.5205e-04 - accuracy: 1.0000 - val_loss: 0.3685 - val_accuracy: 0.8778\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 2.1887e-04 - accuracy: 1.0000 - val_loss: 0.3789 - val_accuracy: 0.8778\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.4751e-04 - accuracy: 1.0000 - val_loss: 0.3872 - val_accuracy: 0.8778\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.0258e-04 - accuracy: 1.0000 - val_loss: 0.3957 - val_accuracy: 0.8778\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 7.4849e-05 - accuracy: 1.0000 - val_loss: 0.4028 - val_accuracy: 0.8833\n",
      "200/200 [==============================] - 0s 273us/step\n",
      "Accuracy on test data for mode binary is 92.00000166893005\n",
      "Model: \"sequential_20\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_58 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_59 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_60 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4448 - accuracy: 0.7889 - val_loss: 0.4217 - val_accuracy: 0.8111\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0316 - accuracy: 0.9926 - val_loss: 0.3346 - val_accuracy: 0.8722\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0033 - accuracy: 1.0000 - val_loss: 0.3532 - val_accuracy: 0.8722\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0013 - accuracy: 1.0000 - val_loss: 0.3629 - val_accuracy: 0.8889\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 7.6222e-04 - accuracy: 1.0000 - val_loss: 0.3742 - val_accuracy: 0.8833\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 4.7196e-04 - accuracy: 1.0000 - val_loss: 0.3852 - val_accuracy: 0.8889\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 2.8422e-04 - accuracy: 1.0000 - val_loss: 0.3918 - val_accuracy: 0.8944\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.8078e-04 - accuracy: 1.0000 - val_loss: 0.3945 - val_accuracy: 0.8833\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 1.0819e-04 - accuracy: 1.0000 - val_loss: 0.4017 - val_accuracy: 0.8778\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 7.0179e-05 - accuracy: 1.0000 - val_loss: 0.4117 - val_accuracy: 0.8778\n",
      "200/200 [==============================] - 0s 271us/step\n",
      "Accuracy on test data for mode binary is 90.49999713897705\n",
      "Model: \"sequential_21\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_61 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_62 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_63 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4526 - accuracy: 0.7840 - val_loss: 0.2985 - val_accuracy: 0.8778\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0286 - accuracy: 0.9963 - val_loss: 0.3818 - val_accuracy: 0.8722\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0025 - accuracy: 1.0000 - val_loss: 0.3349 - val_accuracy: 0.9056\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 9.7693e-04 - accuracy: 1.0000 - val_loss: 0.3493 - val_accuracy: 0.9056\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.3684e-04 - accuracy: 1.0000 - val_loss: 0.3614 - val_accuracy: 0.9167\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.1755e-04 - accuracy: 1.0000 - val_loss: 0.3801 - val_accuracy: 0.9167\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.8105e-04 - accuracy: 1.0000 - val_loss: 0.3930 - val_accuracy: 0.9111\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.1040e-04 - accuracy: 1.0000 - val_loss: 0.4004 - val_accuracy: 0.9167\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 7.3752e-05 - accuracy: 1.0000 - val_loss: 0.4046 - val_accuracy: 0.9167\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.1690e-05 - accuracy: 1.0000 - val_loss: 0.4130 - val_accuracy: 0.9111\n",
      "200/200 [==============================] - 0s 280us/step\n",
      "Accuracy on test data for mode binary is 89.99999761581421\n",
      "Model: \"sequential_22\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_64 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_65 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_66 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4496 - accuracy: 0.7883 - val_loss: 0.3407 - val_accuracy: 0.8611\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0297 - accuracy: 0.9938 - val_loss: 0.3718 - val_accuracy: 0.8611\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0023 - accuracy: 1.0000 - val_loss: 0.3655 - val_accuracy: 0.8722\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.3515e-04 - accuracy: 1.0000 - val_loss: 0.4027 - val_accuracy: 0.8778\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.3367e-04 - accuracy: 1.0000 - val_loss: 0.4222 - val_accuracy: 0.8889\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.2266e-04 - accuracy: 1.0000 - val_loss: 0.4473 - val_accuracy: 0.9000\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 7.1804e-05 - accuracy: 1.0000 - val_loss: 0.4641 - val_accuracy: 0.9000\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.7783e-05 - accuracy: 1.0000 - val_loss: 0.4787 - val_accuracy: 0.8944\n",
      "Epoch 9/10\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1620/1620 [==============================] - 4s 2ms/step - loss: 3.3731e-05 - accuracy: 1.0000 - val_loss: 0.4908 - val_accuracy: 0.8889\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.5023e-05 - accuracy: 1.0000 - val_loss: 0.5022 - val_accuracy: 0.8944\n",
      "200/200 [==============================] - 0s 256us/step\n",
      "Accuracy on test data for mode binary is 91.50000214576721\n",
      "Model: \"sequential_23\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_67 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_68 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_69 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4533 - accuracy: 0.7901 - val_loss: 0.4290 - val_accuracy: 0.8278\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0337 - accuracy: 0.9926 - val_loss: 0.3208 - val_accuracy: 0.8611\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0030 - accuracy: 1.0000 - val_loss: 0.3374 - val_accuracy: 0.8667\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0013 - accuracy: 1.0000 - val_loss: 0.3541 - val_accuracy: 0.8833\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 7.3396e-04 - accuracy: 1.0000 - val_loss: 0.3694 - val_accuracy: 0.8778\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 4.3717e-04 - accuracy: 1.0000 - val_loss: 0.3869 - val_accuracy: 0.8667\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 2.7506e-04 - accuracy: 1.0000 - val_loss: 0.3990 - val_accuracy: 0.8667\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 1.8201e-04 - accuracy: 1.0000 - val_loss: 0.4103 - val_accuracy: 0.8667\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.2578e-04 - accuracy: 1.0000 - val_loss: 0.4212 - val_accuracy: 0.8667\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 9.1255e-05 - accuracy: 1.0000 - val_loss: 0.4302 - val_accuracy: 0.8667\n",
      "200/200 [==============================] - 0s 256us/step\n",
      "Accuracy on test data for mode binary is 91.00000262260437\n",
      "Model: \"sequential_24\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_70 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_71 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_72 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4770 - accuracy: 0.7679 - val_loss: 0.3051 - val_accuracy: 0.8611\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0433 - accuracy: 0.9901 - val_loss: 0.3545 - val_accuracy: 0.8778\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0036 - accuracy: 0.9994 - val_loss: 0.3608 - val_accuracy: 0.8500\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 8.1252e-04 - accuracy: 1.0000 - val_loss: 0.3629 - val_accuracy: 0.8889\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.3268e-04 - accuracy: 1.0000 - val_loss: 0.3798 - val_accuracy: 0.8833\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 2.6509e-04 - accuracy: 1.0000 - val_loss: 0.3961 - val_accuracy: 0.8833\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 1.7392e-04 - accuracy: 1.0000 - val_loss: 0.4085 - val_accuracy: 0.8833\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 1.1752e-04 - accuracy: 1.0000 - val_loss: 0.4187 - val_accuracy: 0.8833\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 8.3333e-05 - accuracy: 1.0000 - val_loss: 0.4285 - val_accuracy: 0.8778\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 6.1317e-05 - accuracy: 1.0000 - val_loss: 0.4361 - val_accuracy: 0.8778\n",
      "200/200 [==============================] - 0s 227us/step\n",
      "Accuracy on test data for mode binary is 91.50000214576721\n",
      "Model: \"sequential_25\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_73 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_74 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_75 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.4665 - accuracy: 0.7710 - val_loss: 0.2984 - val_accuracy: 0.9000\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0367 - accuracy: 0.9914 - val_loss: 0.3533 - val_accuracy: 0.8722\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0044 - accuracy: 1.0000 - val_loss: 0.3101 - val_accuracy: 0.8944\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0013 - accuracy: 1.0000 - val_loss: 0.3203 - val_accuracy: 0.8944\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 7.0388e-04 - accuracy: 1.0000 - val_loss: 0.3294 - val_accuracy: 0.8944\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 4.2670e-04 - accuracy: 1.0000 - val_loss: 0.3364 - val_accuracy: 0.8889\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 2.8414e-04 - accuracy: 1.0000 - val_loss: 0.3444 - val_accuracy: 0.8889\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 3s 2ms/step - loss: 1.9891e-04 - accuracy: 1.0000 - val_loss: 0.3503 - val_accuracy: 0.8889\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 3s 2ms/step - loss: 1.4483e-04 - accuracy: 1.0000 - val_loss: 0.3553 - val_accuracy: 0.8889\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 3s 2ms/step - loss: 1.0883e-04 - accuracy: 1.0000 - val_loss: 0.3602 - val_accuracy: 0.8889\n",
      "200/200 [==============================] - 0s 254us/step\n",
      "Accuracy on test data for mode binary is 91.50000214576721\n",
      "Model: \"sequential_26\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_76 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_77 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_78 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 47s 29ms/step - loss: 0.4623 - accuracy: 0.7698 - val_loss: 0.3412 - val_accuracy: 0.8611\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 7s 4ms/step - loss: 0.0486 - accuracy: 0.9864 - val_loss: 0.4791 - val_accuracy: 0.8556\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 6s 4ms/step - loss: 0.0076 - accuracy: 0.9981 - val_loss: 0.4059 - val_accuracy: 0.8778\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 6s 4ms/step - loss: 0.0017 - accuracy: 1.0000 - val_loss: 0.4253 - val_accuracy: 0.8778\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 6s 4ms/step - loss: 7.8188e-04 - accuracy: 1.0000 - val_loss: 0.4359 - val_accuracy: 0.8778\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 4.6832e-04 - accuracy: 1.0000 - val_loss: 0.4510 - val_accuracy: 0.8778\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 3.0550e-04 - accuracy: 1.0000 - val_loss: 0.4644 - val_accuracy: 0.8778\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 2.0086e-04 - accuracy: 1.0000 - val_loss: 0.4789 - val_accuracy: 0.8778\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.3711e-04 - accuracy: 1.0000 - val_loss: 0.4921 - val_accuracy: 0.8778\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 9.6984e-05 - accuracy: 1.0000 - val_loss: 0.5058 - val_accuracy: 0.8833\n",
      "200/200 [==============================] - 0s 453us/step\n",
      "Accuracy on test data for mode count is 87.5\n",
      "Model: \"sequential_27\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_79 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_80 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_81 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4769 - accuracy: 0.7840 - val_loss: 0.3436 - val_accuracy: 0.8722\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 6s 4ms/step - loss: 0.0533 - accuracy: 0.9864 - val_loss: 0.3978 - val_accuracy: 0.8444\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0047 - accuracy: 1.0000 - val_loss: 0.4448 - val_accuracy: 0.8444\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 9.8244e-04 - accuracy: 1.0000 - val_loss: 0.4689 - val_accuracy: 0.8222\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.6708e-04 - accuracy: 1.0000 - val_loss: 0.4959 - val_accuracy: 0.8278\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.7282e-04 - accuracy: 1.0000 - val_loss: 0.5173 - val_accuracy: 0.8278\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.7817e-04 - accuracy: 1.0000 - val_loss: 0.5333 - val_accuracy: 0.8278\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.2411e-04 - accuracy: 1.0000 - val_loss: 0.5484 - val_accuracy: 0.8278\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 8.9931e-05 - accuracy: 1.0000 - val_loss: 0.5629 - val_accuracy: 0.8222\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.6599e-05 - accuracy: 1.0000 - val_loss: 0.5740 - val_accuracy: 0.8222\n",
      "200/200 [==============================] - 0s 273us/step\n",
      "Accuracy on test data for mode count is 87.5\n",
      "Model: \"sequential_28\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_82 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_83 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_84 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 6s 4ms/step - loss: 0.4790 - accuracy: 0.7636 - val_loss: 0.3411 - val_accuracy: 0.8444\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0509 - accuracy: 0.9877 - val_loss: 0.3552 - val_accuracy: 0.8611\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0051 - accuracy: 1.0000 - val_loss: 0.3801 - val_accuracy: 0.8667\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0018 - accuracy: 1.0000 - val_loss: 0.4087 - val_accuracy: 0.8667\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 9.9357e-04 - accuracy: 1.0000 - val_loss: 0.4302 - val_accuracy: 0.8556\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.2637e-04 - accuracy: 1.0000 - val_loss: 0.4635 - val_accuracy: 0.8611\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.0564e-04 - accuracy: 1.0000 - val_loss: 0.5008 - val_accuracy: 0.8556\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.7050e-04 - accuracy: 1.0000 - val_loss: 0.5328 - val_accuracy: 0.8500\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.9344e-04 - accuracy: 1.0000 - val_loss: 0.5543 - val_accuracy: 0.8444\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.4397e-04 - accuracy: 1.0000 - val_loss: 0.5722 - val_accuracy: 0.8389\n",
      "200/200 [==============================] - 0s 262us/step\n",
      "Accuracy on test data for mode count is 88.99999856948853\n",
      "Model: \"sequential_29\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_85 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_86 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_87 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4874 - accuracy: 0.7531 - val_loss: 0.2906 - val_accuracy: 0.8889\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0452 - accuracy: 0.9895 - val_loss: 0.3268 - val_accuracy: 0.8611\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0034 - accuracy: 1.0000 - val_loss: 0.3912 - val_accuracy: 0.8500\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 9.8738e-04 - accuracy: 1.0000 - val_loss: 0.4178 - val_accuracy: 0.8500\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 5.3952e-04 - accuracy: 1.0000 - val_loss: 0.4461 - val_accuracy: 0.8389\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.3695e-04 - accuracy: 1.0000 - val_loss: 0.4767 - val_accuracy: 0.8389\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.2912e-04 - accuracy: 1.0000 - val_loss: 0.5015 - val_accuracy: 0.8444\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.6303e-04 - accuracy: 1.0000 - val_loss: 0.5290 - val_accuracy: 0.8389\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.2060e-04 - accuracy: 1.0000 - val_loss: 0.5512 - val_accuracy: 0.8389\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 9.1630e-05 - accuracy: 1.0000 - val_loss: 0.5716 - val_accuracy: 0.8333\n",
      "200/200 [==============================] - 0s 260us/step\n",
      "Accuracy on test data for mode count is 86.50000095367432\n",
      "Model: \"sequential_30\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_88 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_89 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_90 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4748 - accuracy: 0.7654 - val_loss: 0.3063 - val_accuracy: 0.8833\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0464 - accuracy: 0.9895 - val_loss: 0.3170 - val_accuracy: 0.8889\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0050 - accuracy: 1.0000 - val_loss: 0.3339 - val_accuracy: 0.8833\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0014 - accuracy: 1.0000 - val_loss: 0.3559 - val_accuracy: 0.8944\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 7.2182e-04 - accuracy: 1.0000 - val_loss: 0.3745 - val_accuracy: 0.8944\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.1963e-04 - accuracy: 1.0000 - val_loss: 0.3865 - val_accuracy: 0.8833\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 2.6600e-04 - accuracy: 1.0000 - val_loss: 0.3970 - val_accuracy: 0.8833\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 1.7522e-04 - accuracy: 1.0000 - val_loss: 0.4053 - val_accuracy: 0.8778\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 1.1462e-04 - accuracy: 1.0000 - val_loss: 0.4168 - val_accuracy: 0.8833\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 8.1155e-05 - accuracy: 1.0000 - val_loss: 0.4237 - val_accuracy: 0.8833\n",
      "200/200 [==============================] - 0s 266us/step\n",
      "Accuracy on test data for mode count is 88.49999904632568\n",
      "Model: \"sequential_31\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_91 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_92 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_93 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4413 - accuracy: 0.8000 - val_loss: 0.3039 - val_accuracy: 0.8667\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0396 - accuracy: 0.9944 - val_loss: 0.3533 - val_accuracy: 0.8833\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0040 - accuracy: 0.9994 - val_loss: 0.3845 - val_accuracy: 0.8556\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0012 - accuracy: 1.0000 - val_loss: 0.4020 - val_accuracy: 0.8500\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.6808e-04 - accuracy: 1.0000 - val_loss: 0.4144 - val_accuracy: 0.8667\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.2820e-04 - accuracy: 1.0000 - val_loss: 0.4268 - val_accuracy: 0.8667\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.8357e-04 - accuracy: 1.0000 - val_loss: 0.4433 - val_accuracy: 0.8667\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.9744e-04 - accuracy: 1.0000 - val_loss: 0.4549 - val_accuracy: 0.8667\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.4201e-04 - accuracy: 1.0000 - val_loss: 0.4688 - val_accuracy: 0.8667\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.0659e-04 - accuracy: 1.0000 - val_loss: 0.4776 - val_accuracy: 0.8667\n",
      "200/200 [==============================] - 0s 272us/step\n",
      "Accuracy on test data for mode count is 86.50000095367432\n",
      "Model: \"sequential_32\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_94 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_95 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_96 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4507 - accuracy: 0.7907 - val_loss: 0.3223 - val_accuracy: 0.8778\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0410 - accuracy: 0.9907 - val_loss: 0.3194 - val_accuracy: 0.8556\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0056 - accuracy: 0.9988 - val_loss: 0.3346 - val_accuracy: 0.8611\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0016 - accuracy: 1.0000 - val_loss: 0.3398 - val_accuracy: 0.8778\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 8.8483e-04 - accuracy: 1.0000 - val_loss: 0.3495 - val_accuracy: 0.8722\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.6046e-04 - accuracy: 1.0000 - val_loss: 0.3569 - val_accuracy: 0.8722\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.7851e-04 - accuracy: 1.0000 - val_loss: 0.3660 - val_accuracy: 0.8611\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 2.6669e-04 - accuracy: 1.0000 - val_loss: 0.3749 - val_accuracy: 0.8611\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.9319e-04 - accuracy: 1.0000 - val_loss: 0.3845 - val_accuracy: 0.8667\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.4375e-04 - accuracy: 1.0000 - val_loss: 0.3957 - val_accuracy: 0.8500\n",
      "200/200 [==============================] - 0s 258us/step\n",
      "Accuracy on test data for mode count is 87.5\n",
      "Model: \"sequential_33\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_97 (Dense)             (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_98 (Dense)             (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_99 (Dense)             (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4680 - accuracy: 0.7599 - val_loss: 0.3178 - val_accuracy: 0.8556\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0348 - accuracy: 0.9932 - val_loss: 0.5373 - val_accuracy: 0.8611\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0038 - accuracy: 1.0000 - val_loss: 0.4943 - val_accuracy: 0.8500\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0013 - accuracy: 1.0000 - val_loss: 0.5133 - val_accuracy: 0.8444\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 7.2941e-04 - accuracy: 1.0000 - val_loss: 0.5326 - val_accuracy: 0.8389\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 4.5010e-04 - accuracy: 1.0000 - val_loss: 0.5533 - val_accuracy: 0.8389\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.9954e-04 - accuracy: 1.0000 - val_loss: 0.5669 - val_accuracy: 0.8444\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.0455e-04 - accuracy: 1.0000 - val_loss: 0.5812 - val_accuracy: 0.8389\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.4413e-04 - accuracy: 1.0000 - val_loss: 0.5895 - val_accuracy: 0.8389\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.0373e-04 - accuracy: 1.0000 - val_loss: 0.5997 - val_accuracy: 0.8444\n",
      "200/200 [==============================] - 0s 270us/step\n",
      "Accuracy on test data for mode count is 88.99999856948853\n",
      "Model: \"sequential_34\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_100 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_101 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_102 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4640 - accuracy: 0.7877 - val_loss: 0.3354 - val_accuracy: 0.8722\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0408 - accuracy: 0.9895 - val_loss: 0.4106 - val_accuracy: 0.8778\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0036 - accuracy: 1.0000 - val_loss: 0.4057 - val_accuracy: 0.8722\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.8988e-04 - accuracy: 1.0000 - val_loss: 0.4786 - val_accuracy: 0.8778\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.9923e-04 - accuracy: 1.0000 - val_loss: 0.5245 - val_accuracy: 0.8778\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.8226e-04 - accuracy: 1.0000 - val_loss: 0.5413 - val_accuracy: 0.8833\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.1953e-04 - accuracy: 1.0000 - val_loss: 0.5506 - val_accuracy: 0.8667\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 8.2697e-05 - accuracy: 1.0000 - val_loss: 0.5621 - val_accuracy: 0.8667\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.9750e-05 - accuracy: 1.0000 - val_loss: 0.5752 - val_accuracy: 0.8611\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.5338e-05 - accuracy: 1.0000 - val_loss: 0.5886 - val_accuracy: 0.8611\n",
      "200/200 [==============================] - 0s 252us/step\n",
      "Accuracy on test data for mode count is 88.99999856948853\n",
      "Model: \"sequential_35\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_103 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_104 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_105 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4691 - accuracy: 0.7858 - val_loss: 0.4152 - val_accuracy: 0.8278\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 0.0452 - accuracy: 0.9907 - val_loss: 0.3874 - val_accuracy: 0.8556\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0039 - accuracy: 1.0000 - val_loss: 0.4216 - val_accuracy: 0.8556\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 8.6867e-04 - accuracy: 1.0000 - val_loss: 0.4903 - val_accuracy: 0.8444\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.1110e-04 - accuracy: 1.0000 - val_loss: 0.5245 - val_accuracy: 0.8444\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.6876e-04 - accuracy: 1.0000 - val_loss: 0.5502 - val_accuracy: 0.8444\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 1.0577e-04 - accuracy: 1.0000 - val_loss: 0.5627 - val_accuracy: 0.8444\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 6.9615e-05 - accuracy: 1.0000 - val_loss: 0.5762 - val_accuracy: 0.8444\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 4.9203e-05 - accuracy: 1.0000 - val_loss: 0.5859 - val_accuracy: 0.8444\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 3.5985e-05 - accuracy: 1.0000 - val_loss: 0.5965 - val_accuracy: 0.8444\n",
      "200/200 [==============================] - 0s 247us/step\n",
      "Accuracy on test data for mode count is 89.49999809265137\n",
      "Model: \"sequential_36\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_106 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_107 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_108 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.5007 - accuracy: 0.7562 - val_loss: 0.3489 - val_accuracy: 0.8556\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0254 - accuracy: 0.9963 - val_loss: 0.4399 - val_accuracy: 0.8611\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0019 - accuracy: 1.0000 - val_loss: 0.4727 - val_accuracy: 0.8556\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 8.9476e-04 - accuracy: 1.0000 - val_loss: 0.5030 - val_accuracy: 0.8556\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.5623e-04 - accuracy: 1.0000 - val_loss: 0.5448 - val_accuracy: 0.8611\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.5365e-04 - accuracy: 1.0000 - val_loss: 0.5885 - val_accuracy: 0.8611\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.5925e-04 - accuracy: 1.0000 - val_loss: 0.6206 - val_accuracy: 0.8611\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.0854e-04 - accuracy: 1.0000 - val_loss: 0.6415 - val_accuracy: 0.8611\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 7.5519e-05 - accuracy: 1.0000 - val_loss: 0.6608 - val_accuracy: 0.8611\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.4388e-05 - accuracy: 1.0000 - val_loss: 0.6749 - val_accuracy: 0.8611\n",
      "200/200 [==============================] - 0s 445us/step\n",
      "Accuracy on test data for mode tfidf is 86.00000143051147\n",
      "Model: \"sequential_37\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_109 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_110 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_111 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4797 - accuracy: 0.7747 - val_loss: 0.3444 - val_accuracy: 0.8444\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0258 - accuracy: 0.9944 - val_loss: 0.3929 - val_accuracy: 0.8500\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0022 - accuracy: 1.0000 - val_loss: 0.4289 - val_accuracy: 0.8500\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 9.7918e-04 - accuracy: 1.0000 - val_loss: 0.4488 - val_accuracy: 0.8500\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.5539e-04 - accuracy: 1.0000 - val_loss: 0.4692 - val_accuracy: 0.8444\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.6102e-04 - accuracy: 1.0000 - val_loss: 0.4888 - val_accuracy: 0.8500\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 2.4476e-04 - accuracy: 1.0000 - val_loss: 0.5158 - val_accuracy: 0.8444\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.6690e-04 - accuracy: 1.0000 - val_loss: 0.5426 - val_accuracy: 0.8389\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.1821e-04 - accuracy: 1.0000 - val_loss: 0.5597 - val_accuracy: 0.8389\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 8.7835e-05 - accuracy: 1.0000 - val_loss: 0.5741 - val_accuracy: 0.8389\n",
      "200/200 [==============================] - 0s 271us/step\n",
      "Accuracy on test data for mode tfidf is 83.49999785423279\n",
      "Model: \"sequential_38\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_112 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_113 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_114 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.5166 - accuracy: 0.7321 - val_loss: 0.3438 - val_accuracy: 0.8444\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0462 - accuracy: 0.9883 - val_loss: 0.6416 - val_accuracy: 0.8333\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0025 - accuracy: 1.0000 - val_loss: 0.4212 - val_accuracy: 0.8500\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.3989e-04 - accuracy: 1.0000 - val_loss: 0.4434 - val_accuracy: 0.8500\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.8941e-04 - accuracy: 1.0000 - val_loss: 0.4544 - val_accuracy: 0.8444\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.8013e-04 - accuracy: 1.0000 - val_loss: 0.4681 - val_accuracy: 0.8500\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.2104e-04 - accuracy: 1.0000 - val_loss: 0.4831 - val_accuracy: 0.8500\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 8.5044e-05 - accuracy: 1.0000 - val_loss: 0.4954 - val_accuracy: 0.8444\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.3030e-05 - accuracy: 1.0000 - val_loss: 0.5077 - val_accuracy: 0.8444\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 4.8216e-05 - accuracy: 1.0000 - val_loss: 0.5190 - val_accuracy: 0.8500\n",
      "200/200 [==============================] - 0s 267us/step\n",
      "Accuracy on test data for mode tfidf is 83.99999737739563\n",
      "Model: \"sequential_39\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_115 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_116 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_117 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4757 - accuracy: 0.7747 - val_loss: 0.4027 - val_accuracy: 0.8167\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0301 - accuracy: 0.9932 - val_loss: 0.4330 - val_accuracy: 0.8611\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0016 - accuracy: 1.0000 - val_loss: 0.4775 - val_accuracy: 0.8667\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.4432e-04 - accuracy: 1.0000 - val_loss: 0.4993 - val_accuracy: 0.8667\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.9699e-04 - accuracy: 1.0000 - val_loss: 0.5147 - val_accuracy: 0.8722\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.5598e-04 - accuracy: 1.0000 - val_loss: 0.5293 - val_accuracy: 0.8722\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.7843e-04 - accuracy: 1.0000 - val_loss: 0.5411 - val_accuracy: 0.8722\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.3013e-04 - accuracy: 1.0000 - val_loss: 0.5548 - val_accuracy: 0.8722\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 9.8220e-05 - accuracy: 1.0000 - val_loss: 0.5662 - val_accuracy: 0.8722\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 7.6099e-05 - accuracy: 1.0000 - val_loss: 0.5757 - val_accuracy: 0.8722\n",
      "200/200 [==============================] - 0s 345us/step\n",
      "Accuracy on test data for mode tfidf is 87.00000047683716\n",
      "Model: \"sequential_40\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_118 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_119 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_120 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 6s 3ms/step - loss: 0.4881 - accuracy: 0.7623 - val_loss: 0.3670 - val_accuracy: 0.8333\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0268 - accuracy: 0.9938 - val_loss: 0.4283 - val_accuracy: 0.8444\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0021 - accuracy: 1.0000 - val_loss: 0.4576 - val_accuracy: 0.8500\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 9.2085e-04 - accuracy: 1.0000 - val_loss: 0.4747 - val_accuracy: 0.8556\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 5.3358e-04 - accuracy: 1.0000 - val_loss: 0.4915 - val_accuracy: 0.8556\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 3.5281e-04 - accuracy: 1.0000 - val_loss: 0.5126 - val_accuracy: 0.8611\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.4430e-04 - accuracy: 1.0000 - val_loss: 0.5253 - val_accuracy: 0.8556\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.7325e-04 - accuracy: 1.0000 - val_loss: 0.5402 - val_accuracy: 0.8556\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.3008e-04 - accuracy: 1.0000 - val_loss: 0.5545 - val_accuracy: 0.8556\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 6s 3ms/step - loss: 1.0052e-04 - accuracy: 1.0000 - val_loss: 0.5647 - val_accuracy: 0.8556\n",
      "200/200 [==============================] - 0s 421us/step\n",
      "Accuracy on test data for mode tfidf is 86.00000143051147\n",
      "Model: \"sequential_41\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_121 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_122 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_123 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4818 - accuracy: 0.7728 - val_loss: 0.3774 - val_accuracy: 0.8500\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0245 - accuracy: 0.9957 - val_loss: 0.4388 - val_accuracy: 0.8500\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0020 - accuracy: 1.0000 - val_loss: 0.4830 - val_accuracy: 0.8444\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 9.2735e-04 - accuracy: 1.0000 - val_loss: 0.5109 - val_accuracy: 0.8389\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.6530e-04 - accuracy: 1.0000 - val_loss: 0.5320 - val_accuracy: 0.8389\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 3.7102e-04 - accuracy: 1.0000 - val_loss: 0.5520 - val_accuracy: 0.8389\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.5975e-04 - accuracy: 1.0000 - val_loss: 0.5708 - val_accuracy: 0.8389\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.9108e-04 - accuracy: 1.0000 - val_loss: 0.5908 - val_accuracy: 0.8389\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.4286e-04 - accuracy: 1.0000 - val_loss: 0.6126 - val_accuracy: 0.8333\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.1072e-04 - accuracy: 1.0000 - val_loss: 0.6322 - val_accuracy: 0.8333\n",
      "200/200 [==============================] - 0s 349us/step\n",
      "Accuracy on test data for mode tfidf is 87.5\n",
      "Model: \"sequential_42\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_124 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_125 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_126 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4871 - accuracy: 0.7827 - val_loss: 0.3568 - val_accuracy: 0.8611\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0354 - accuracy: 0.9901 - val_loss: 0.4035 - val_accuracy: 0.8778\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0033 - accuracy: 0.9994 - val_loss: 0.4346 - val_accuracy: 0.8833\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0018 - accuracy: 0.9994 - val_loss: 0.4552 - val_accuracy: 0.8833\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 9.7715e-04 - accuracy: 0.9994 - val_loss: 0.4723 - val_accuracy: 0.8667\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 7.6489e-04 - accuracy: 1.0000 - val_loss: 0.4858 - val_accuracy: 0.8667\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.4607e-04 - accuracy: 1.0000 - val_loss: 0.5017 - val_accuracy: 0.8667\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 5.6199e-04 - accuracy: 1.0000 - val_loss: 0.5148 - val_accuracy: 0.8667\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.2258e-04 - accuracy: 1.0000 - val_loss: 0.5245 - val_accuracy: 0.8667\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 8.9454e-05 - accuracy: 1.0000 - val_loss: 0.5319 - val_accuracy: 0.8667\n",
      "200/200 [==============================] - 0s 261us/step\n",
      "Accuracy on test data for mode tfidf is 86.00000143051147\n",
      "Model: \"sequential_43\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_127 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_128 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_129 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4991 - accuracy: 0.7463 - val_loss: 0.3348 - val_accuracy: 0.8556\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0345 - accuracy: 0.9920 - val_loss: 0.3727 - val_accuracy: 0.8722\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0020 - accuracy: 1.0000 - val_loss: 0.4352 - val_accuracy: 0.8722\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 5.1191e-04 - accuracy: 1.0000 - val_loss: 0.4576 - val_accuracy: 0.8722\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 2.7228e-04 - accuracy: 1.0000 - val_loss: 0.4760 - val_accuracy: 0.8722\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.7178e-04 - accuracy: 1.0000 - val_loss: 0.4896 - val_accuracy: 0.8778\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 1.2002e-04 - accuracy: 1.0000 - val_loss: 0.5044 - val_accuracy: 0.8833\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 8.7814e-05 - accuracy: 1.0000 - val_loss: 0.5182 - val_accuracy: 0.8833\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 2ms/step - loss: 6.6160e-05 - accuracy: 1.0000 - val_loss: 0.5340 - val_accuracy: 0.8778\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.1447e-05 - accuracy: 1.0000 - val_loss: 0.5460 - val_accuracy: 0.8778\n",
      "200/200 [==============================] - 0s 278us/step\n",
      "Accuracy on test data for mode tfidf is 85.50000190734863\n",
      "Model: \"sequential_44\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_130 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_131 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_132 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.4929 - accuracy: 0.7667 - val_loss: 0.3792 - val_accuracy: 0.8333\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0357 - accuracy: 0.9895 - val_loss: 0.5624 - val_accuracy: 0.8333\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0025 - accuracy: 1.0000 - val_loss: 0.5699 - val_accuracy: 0.8333\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 8.4640e-04 - accuracy: 1.0000 - val_loss: 0.5668 - val_accuracy: 0.8222\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 4.0883e-04 - accuracy: 1.0000 - val_loss: 0.5868 - val_accuracy: 0.8333\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 2.3685e-04 - accuracy: 1.0000 - val_loss: 0.6096 - val_accuracy: 0.8333\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.5788e-04 - accuracy: 1.0000 - val_loss: 0.6254 - val_accuracy: 0.8333\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.0999e-04 - accuracy: 1.0000 - val_loss: 0.6423 - val_accuracy: 0.8333\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 7.9824e-05 - accuracy: 1.0000 - val_loss: 0.6562 - val_accuracy: 0.8278\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 6.0566e-05 - accuracy: 1.0000 - val_loss: 0.6786 - val_accuracy: 0.8278\n",
      "200/200 [==============================] - 0s 269us/step\n",
      "Accuracy on test data for mode tfidf is 86.00000143051147\n",
      "Model: \"sequential_45\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_133 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_134 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_135 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4988 - accuracy: 0.7617 - val_loss: 0.3440 - val_accuracy: 0.8667\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0320 - accuracy: 0.9944 - val_loss: 0.4181 - val_accuracy: 0.8611\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0028 - accuracy: 1.0000 - val_loss: 0.4522 - val_accuracy: 0.8667\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0010 - accuracy: 1.0000 - val_loss: 0.4750 - val_accuracy: 0.8444\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 5.5421e-04 - accuracy: 1.0000 - val_loss: 0.5112 - val_accuracy: 0.8333\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 3.0128e-04 - accuracy: 1.0000 - val_loss: 0.5551 - val_accuracy: 0.8333\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.9436e-04 - accuracy: 1.0000 - val_loss: 0.5891 - val_accuracy: 0.8389\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 1.3482e-04 - accuracy: 1.0000 - val_loss: 0.6125 - val_accuracy: 0.8389\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 9.6006e-05 - accuracy: 1.0000 - val_loss: 0.6359 - val_accuracy: 0.8444\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 7.1825e-05 - accuracy: 1.0000 - val_loss: 0.6545 - val_accuracy: 0.8444\n",
      "200/200 [==============================] - 0s 527us/step\n",
      "Accuracy on test data for mode tfidf is 85.50000190734863\n",
      "Model: \"sequential_46\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_136 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_137 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_138 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 6s 4ms/step - loss: 0.6932 - accuracy: 0.4988 - val_loss: 0.6934 - val_accuracy: 0.4556\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6925 - accuracy: 0.5463 - val_loss: 0.6898 - val_accuracy: 0.7222\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6433 - accuracy: 0.7296 - val_loss: 0.5869 - val_accuracy: 0.8056\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.4172 - accuracy: 0.9438 - val_loss: 0.4198 - val_accuracy: 0.8556\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.2045 - accuracy: 0.9685 - val_loss: 0.3557 - val_accuracy: 0.8722\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.1020 - accuracy: 0.9926 - val_loss: 0.2983 - val_accuracy: 0.8722\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0547 - accuracy: 0.9981 - val_loss: 0.2818 - val_accuracy: 0.8611\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 6s 3ms/step - loss: 0.0316 - accuracy: 1.0000 - val_loss: 0.2754 - val_accuracy: 0.8667\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0200 - accuracy: 1.0000 - val_loss: 0.2737 - val_accuracy: 0.8667\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0133 - accuracy: 1.0000 - val_loss: 0.2730 - val_accuracy: 0.8611\n",
      "200/200 [==============================] - 0s 306us/step\n",
      "Accuracy on test data for mode freq is 91.00000262260437\n",
      "Model: \"sequential_47\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_139 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_140 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_141 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6915 - accuracy: 0.5525 - val_loss: 0.6856 - val_accuracy: 0.8000\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6394 - accuracy: 0.8037 - val_loss: 0.5771 - val_accuracy: 0.7833\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.3990 - accuracy: 0.9296 - val_loss: 0.3960 - val_accuracy: 0.8222\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.1757 - accuracy: 0.9778 - val_loss: 0.3173 - val_accuracy: 0.8500\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0781 - accuracy: 0.9951 - val_loss: 0.2904 - val_accuracy: 0.8667\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0379 - accuracy: 1.0000 - val_loss: 0.2768 - val_accuracy: 0.8722\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0211 - accuracy: 1.0000 - val_loss: 0.2758 - val_accuracy: 0.8667\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0128 - accuracy: 1.0000 - val_loss: 0.2740 - val_accuracy: 0.8722\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0083 - accuracy: 1.0000 - val_loss: 0.2755 - val_accuracy: 0.8667\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0059 - accuracy: 1.0000 - val_loss: 0.2789 - val_accuracy: 0.8722\n",
      "200/200 [==============================] - 0s 254us/step\n",
      "Accuracy on test data for mode freq is 91.00000262260437\n",
      "Model: \"sequential_48\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_142 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_143 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_144 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.6904 - accuracy: 0.5630 - val_loss: 0.6796 - val_accuracy: 0.7722\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.6242 - accuracy: 0.7654 - val_loss: 0.5730 - val_accuracy: 0.8333\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.3732 - accuracy: 0.9395 - val_loss: 0.3928 - val_accuracy: 0.8556\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.1649 - accuracy: 0.9784 - val_loss: 0.3092 - val_accuracy: 0.8722\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0730 - accuracy: 0.9957 - val_loss: 0.2820 - val_accuracy: 0.8833\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0352 - accuracy: 1.0000 - val_loss: 0.2740 - val_accuracy: 0.8722\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0189 - accuracy: 1.0000 - val_loss: 0.2753 - val_accuracy: 0.8722\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0116 - accuracy: 1.0000 - val_loss: 0.2730 - val_accuracy: 0.8722\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0075 - accuracy: 1.0000 - val_loss: 0.2751 - val_accuracy: 0.8722\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0051 - accuracy: 1.0000 - val_loss: 0.2791 - val_accuracy: 0.8667\n",
      "200/200 [==============================] - 0s 259us/step\n",
      "Accuracy on test data for mode freq is 91.50000214576721\n",
      "Model: \"sequential_49\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_145 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_146 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_147 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6909 - accuracy: 0.5117 - val_loss: 0.6875 - val_accuracy: 0.4556\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.6354 - accuracy: 0.6383 - val_loss: 0.6027 - val_accuracy: 0.7444\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.3986 - accuracy: 0.9136 - val_loss: 0.3888 - val_accuracy: 0.8611\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.1351 - accuracy: 0.9821 - val_loss: 0.3017 - val_accuracy: 0.8667\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0474 - accuracy: 0.9981 - val_loss: 0.2890 - val_accuracy: 0.8778\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0207 - accuracy: 1.0000 - val_loss: 0.2850 - val_accuracy: 0.8722\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0111 - accuracy: 1.0000 - val_loss: 0.2820 - val_accuracy: 0.8667\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0069 - accuracy: 1.0000 - val_loss: 0.2874 - val_accuracy: 0.8667\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0046 - accuracy: 1.0000 - val_loss: 0.2930 - val_accuracy: 0.8667\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0033 - accuracy: 1.0000 - val_loss: 0.3032 - val_accuracy: 0.8722\n",
      "200/200 [==============================] - 0s 281us/step\n",
      "Accuracy on test data for mode freq is 88.99999856948853\n",
      "Model: \"sequential_50\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_148 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_149 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_150 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6909 - accuracy: 0.5630 - val_loss: 0.6893 - val_accuracy: 0.4556\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.6188 - accuracy: 0.7654 - val_loss: 0.5405 - val_accuracy: 0.8000\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.3328 - accuracy: 0.9444 - val_loss: 0.3637 - val_accuracy: 0.8444\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.1317 - accuracy: 0.9840 - val_loss: 0.2985 - val_accuracy: 0.8778\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0559 - accuracy: 0.9975 - val_loss: 0.2848 - val_accuracy: 0.8500\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0269 - accuracy: 1.0000 - val_loss: 0.2912 - val_accuracy: 0.8667\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0148 - accuracy: 1.0000 - val_loss: 0.2763 - val_accuracy: 0.8611\n",
      "Epoch 8/10\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0091 - accuracy: 1.0000 - val_loss: 0.2812 - val_accuracy: 0.8611\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0062 - accuracy: 1.0000 - val_loss: 0.2814 - val_accuracy: 0.8667\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0043 - accuracy: 1.0000 - val_loss: 0.2848 - val_accuracy: 0.8667\n",
      "200/200 [==============================] - 0s 267us/step\n",
      "Accuracy on test data for mode freq is 91.50000214576721\n",
      "Model: \"sequential_51\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_151 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_152 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_153 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 6s 3ms/step - loss: 0.6917 - accuracy: 0.5401 - val_loss: 0.6861 - val_accuracy: 0.8167\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6392 - accuracy: 0.7833 - val_loss: 0.5713 - val_accuracy: 0.8000\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.3740 - accuracy: 0.9352 - val_loss: 0.3739 - val_accuracy: 0.8333\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.1473 - accuracy: 0.9796 - val_loss: 0.3140 - val_accuracy: 0.8500\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0624 - accuracy: 0.9969 - val_loss: 0.2832 - val_accuracy: 0.8778\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0293 - accuracy: 1.0000 - val_loss: 0.2850 - val_accuracy: 0.8667\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0158 - accuracy: 1.0000 - val_loss: 0.2818 - val_accuracy: 0.8611\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0096 - accuracy: 1.0000 - val_loss: 0.2783 - val_accuracy: 0.8722\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0063 - accuracy: 1.0000 - val_loss: 0.2825 - val_accuracy: 0.8667\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0043 - accuracy: 1.0000 - val_loss: 0.2874 - val_accuracy: 0.8722\n",
      "200/200 [==============================] - 0s 250us/step\n",
      "Accuracy on test data for mode freq is 90.49999713897705\n",
      "Model: \"sequential_52\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_154 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_155 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_156 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6908 - accuracy: 0.5605 - val_loss: 0.6830 - val_accuracy: 0.5222\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6122 - accuracy: 0.7506 - val_loss: 0.5633 - val_accuracy: 0.7778\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.3325 - accuracy: 0.9488 - val_loss: 0.3608 - val_accuracy: 0.8833\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.1239 - accuracy: 0.9889 - val_loss: 0.3038 - val_accuracy: 0.8667\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0507 - accuracy: 0.9988 - val_loss: 0.3170 - val_accuracy: 0.8722\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0238 - accuracy: 1.0000 - val_loss: 0.2845 - val_accuracy: 0.8722\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0131 - accuracy: 1.0000 - val_loss: 0.2807 - val_accuracy: 0.8556\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0083 - accuracy: 1.0000 - val_loss: 0.2846 - val_accuracy: 0.8611\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0056 - accuracy: 1.0000 - val_loss: 0.2871 - val_accuracy: 0.8611\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0040 - accuracy: 1.0000 - val_loss: 0.2946 - val_accuracy: 0.8667\n",
      "200/200 [==============================] - 0s 263us/step\n",
      "Accuracy on test data for mode freq is 89.49999809265137\n",
      "Model: \"sequential_53\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_157 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_158 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_159 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6877 - accuracy: 0.5907 - val_loss: 0.6727 - val_accuracy: 0.6833\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.5731 - accuracy: 0.8599 - val_loss: 0.4930 - val_accuracy: 0.8111\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.2756 - accuracy: 0.9506 - val_loss: 0.3400 - val_accuracy: 0.8722\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.1072 - accuracy: 0.9926 - val_loss: 0.2908 - val_accuracy: 0.8667\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0439 - accuracy: 0.9981 - val_loss: 0.2925 - val_accuracy: 0.8833\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0216 - accuracy: 1.0000 - val_loss: 0.2752 - val_accuracy: 0.8667\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0125 - accuracy: 1.0000 - val_loss: 0.2775 - val_accuracy: 0.8667\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0078 - accuracy: 1.0000 - val_loss: 0.2776 - val_accuracy: 0.8611\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0053 - accuracy: 1.0000 - val_loss: 0.2864 - val_accuracy: 0.8778\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0038 - accuracy: 1.0000 - val_loss: 0.2864 - val_accuracy: 0.8778\n",
      "200/200 [==============================] - 0s 270us/step\n",
      "Accuracy on test data for mode freq is 90.49999713897705\n",
      "Model: \"sequential_54\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_160 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_161 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_162 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6902 - accuracy: 0.5593 - val_loss: 0.6812 - val_accuracy: 0.8056\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.6126 - accuracy: 0.8414 - val_loss: 0.5533 - val_accuracy: 0.8167\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.3639 - accuracy: 0.9451 - val_loss: 0.3913 - val_accuracy: 0.8444\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.1635 - accuracy: 0.9809 - val_loss: 0.3160 - val_accuracy: 0.8667\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0725 - accuracy: 0.9957 - val_loss: 0.2931 - val_accuracy: 0.8556\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0317 - accuracy: 1.0000 - val_loss: 0.2846 - val_accuracy: 0.8611\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0152 - accuracy: 1.0000 - val_loss: 0.2802 - val_accuracy: 0.8611\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0086 - accuracy: 1.0000 - val_loss: 0.2812 - val_accuracy: 0.8500\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0053 - accuracy: 1.0000 - val_loss: 0.2852 - val_accuracy: 0.8556\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0036 - accuracy: 1.0000 - val_loss: 0.2900 - val_accuracy: 0.8611\n",
      "200/200 [==============================] - 0s 262us/step\n",
      "Accuracy on test data for mode freq is 90.49999713897705\n",
      "Model: \"sequential_55\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_163 (Dense)            (None, 50)                1288450   \n",
      "_________________________________________________________________\n",
      "dense_164 (Dense)            (None, 25)                1275      \n",
      "_________________________________________________________________\n",
      "dense_165 (Dense)            (None, 1)                 26        \n",
      "=================================================================\n",
      "Total params: 1,289,751\n",
      "Trainable params: 1,289,751\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1620 samples, validate on 180 samples\n",
      "Epoch 1/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.6880 - accuracy: 0.6068 - val_loss: 0.6683 - val_accuracy: 0.6556\n",
      "Epoch 2/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.5577 - accuracy: 0.8741 - val_loss: 0.4666 - val_accuracy: 0.8389\n",
      "Epoch 3/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.2647 - accuracy: 0.9512 - val_loss: 0.3334 - val_accuracy: 0.8778\n",
      "Epoch 4/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.1051 - accuracy: 0.9883 - val_loss: 0.2990 - val_accuracy: 0.8667\n",
      "Epoch 5/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0441 - accuracy: 0.9981 - val_loss: 0.2792 - val_accuracy: 0.8722\n",
      "Epoch 6/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0215 - accuracy: 1.0000 - val_loss: 0.2756 - val_accuracy: 0.8833\n",
      "Epoch 7/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0122 - accuracy: 1.0000 - val_loss: 0.2761 - val_accuracy: 0.8722\n",
      "Epoch 8/10\n",
      "1620/1620 [==============================] - 5s 3ms/step - loss: 0.0077 - accuracy: 1.0000 - val_loss: 0.2787 - val_accuracy: 0.8722\n",
      "Epoch 9/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0052 - accuracy: 1.0000 - val_loss: 0.2850 - val_accuracy: 0.8722\n",
      "Epoch 10/10\n",
      "1620/1620 [==============================] - 4s 3ms/step - loss: 0.0036 - accuracy: 1.0000 - val_loss: 0.2869 - val_accuracy: 0.8667\n",
      "200/200 [==============================] - 0s 271us/step\n",
      "Accuracy on test data for mode freq is 91.00000262260437\n"
     ]
    }
   ],
   "source": [
    "modes = ['binary', 'count', 'tfidf', 'freq']\n",
    "results = pd.DataFrame()\n",
    "for mode in modes:\n",
    "    xtrain = tokenizer.texts_to_matrix(train_reviews, mode=mode)\n",
    "    xtest = tokenizer.texts_to_matrix(test_reviews, mode=mode)\n",
    "    scores = []\n",
    "    for _ in range(10):\n",
    "        model = Model(n_words)\n",
    "        model.fit(xtrain, y_train, epochs=10, batch_size = 8, validation_split=0.1, shuffle=True)\n",
    "        loss, acc = model.evaluate(xtest, y_test)\n",
    "        print(\"Accuracy on test data for mode\", mode, \"is\", acc*100)\n",
    "        scores.append(acc)\n",
    "    results[mode] = scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-24T00:12:04.338845Z",
     "start_time": "2020-03-24T00:12:04.319242Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>binary</th>\n",
       "      <th>count</th>\n",
       "      <th>tfidf</th>\n",
       "      <th>freq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.915</td>\n",
       "      <td>0.875</td>\n",
       "      <td>0.860</td>\n",
       "      <td>0.910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.905</td>\n",
       "      <td>0.875</td>\n",
       "      <td>0.835</td>\n",
       "      <td>0.910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.895</td>\n",
       "      <td>0.890</td>\n",
       "      <td>0.840</td>\n",
       "      <td>0.915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.920</td>\n",
       "      <td>0.865</td>\n",
       "      <td>0.870</td>\n",
       "      <td>0.890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.905</td>\n",
       "      <td>0.885</td>\n",
       "      <td>0.860</td>\n",
       "      <td>0.915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.900</td>\n",
       "      <td>0.865</td>\n",
       "      <td>0.875</td>\n",
       "      <td>0.905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.915</td>\n",
       "      <td>0.875</td>\n",
       "      <td>0.860</td>\n",
       "      <td>0.895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.910</td>\n",
       "      <td>0.890</td>\n",
       "      <td>0.855</td>\n",
       "      <td>0.905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.915</td>\n",
       "      <td>0.890</td>\n",
       "      <td>0.860</td>\n",
       "      <td>0.905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.915</td>\n",
       "      <td>0.895</td>\n",
       "      <td>0.855</td>\n",
       "      <td>0.910</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   binary  count  tfidf   freq\n",
       "0   0.915  0.875  0.860  0.910\n",
       "1   0.905  0.875  0.835  0.910\n",
       "2   0.895  0.890  0.840  0.915\n",
       "3   0.920  0.865  0.870  0.890\n",
       "4   0.905  0.885  0.860  0.915\n",
       "5   0.900  0.865  0.875  0.905\n",
       "6   0.915  0.875  0.860  0.895\n",
       "7   0.910  0.890  0.855  0.905\n",
       "8   0.915  0.890  0.860  0.905\n",
       "9   0.915  0.895  0.855  0.910"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-03-24T00:16:36.326742Z",
     "start_time": "2020-03-24T00:16:36.060194Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "results.boxplot()\n",
    "plt.savefig('boxplot of different word scoring methods.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Further modifications\n",
    "\n",
    "- try to change architecture of model or hyperparameters\n",
    "- try using regularizations to avoid overfitting(3rd 4th epoch is getting 1.0 training accuracy which is not good)\n",
    "- try using ngrams\n",
    "- try new vocab\n",
    "- try data cleaning\n",
    "- try ensemble\n",
    "\n",
    "> there are endless possibilities."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.10"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}