Naive_Bayes_Implementation.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Naive_Bayes_Implementation.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tNGBMrKizCMt",
        "colab_type": "text"
      },
      "source": [
        "# Naïve Bayes Implementation\n",
        "\n",
        "Oscar Contreras Carrasco\n",
        "\n",
        "We are now going to implement Naïve Bayes in Python! For our implementation, we are going to use the MNIST dataset of handwritten numbers. I hope you find this material useful. Feel free to approach me in case you have any questions or comments.\n",
        "\n",
        "Let's first import our libraries\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ArpzyzmHy-ra",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "from tensorflow import keras\n",
        "from sklearn.datasets import load_iris, load_digits\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import accuracy_score\n",
        "import cv2"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WfwJJ2BLHmIP",
        "colab_type": "text"
      },
      "source": [
        "Now we are going  to prepare our dataset. The Keras library already ships with MNIST and we can easily make use of it. In general, the dataset has 60,000 images for training and 10,000 for testing.\n",
        "\n",
        "However, as part of our pre-processing, we are also applying HOG descriptors to our images. This will enhance prediction performance and will reduce the number of total features to evaluate. You can find out more about HOG features using OpenCV at https://docs.opencv.org/3.4.6/d5/d33/structcv_1_1HOGDescriptor.html"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-e4fIAXw8WmD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "dataset = keras.datasets.mnist\n",
        "\n",
        "(X_raw_train, y_train), (X_raw_test, y_test) = dataset.load_data()\n",
        "\n",
        "X_train = np.zeros((X_raw_train.shape[0], 108))\n",
        "X_test = np.zeros((X_raw_test.shape[0], 108))\n",
        "\n",
        "hog = cv2.HOGDescriptor((28, 28), (14, 14), (7, 7), (14, 14), 12)\n",
        "\n",
        "for n in range(X_raw_train.shape[0]):\n",
        "    X_train[n] = hog.compute(X_raw_train[n]).reshape(1, -1)\n",
        "    \n",
        "for n in range(X_raw_test.shape[0]):\n",
        "    X_test[n] = hog.compute(X_raw_test[n]).reshape(1, -1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X-5P8ncsIspx",
        "colab_type": "text"
      },
      "source": [
        "Now let's display some images from our dataset! All images are grayscale and 28x28"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n6gTNU0LxscW",
        "colab_type": "code",
        "outputId": "4c207ec0-552d-4890-af76-211bb57bb155",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 582
        }
      },
      "source": [
        "_, ax = plt.subplots(8, 8, figsize=(10, 10))\n",
        "index = 0\n",
        "\n",
        "for i in range(8):\n",
        "    for j in range(8):\n",
        "        cell = ax[i][j]\n",
        "        cell.set_xticks([])\n",
        "        cell.set_yticks([])\n",
        "        cell.grid(False)\n",
        "        cell.imshow(X_raw_test[index], cmap='gray')\n",
        "        index += 1\n",
        "        \n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkEAAAI1CAYAAADRty56AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZmZmlkj/B/0SHuqRDF4fAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 720x720 with 64 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MmG9_YGnI-xA",
        "colab_type": "text"
      },
      "source": [
        "Now let's calculate the prior probability $p(\\mathbf y)$. This will actually be a vector containing the probabilities of occurrence for each class in our dataset. That is, the numbers from 0 to 9."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YqgPuNzO-psM",
        "colab_type": "code",
        "outputId": "dfb21c71-8124-4344-c308-5adb88ed9e55",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "_, prob_y = np.unique(y_train, return_counts=True)\n",
        "prob_y = prob_y / len(y_train)\n",
        "num_classes = len(prob_y)\n",
        "prob_y"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.09871667, 0.11236667, 0.0993    , 0.10218333, 0.09736667,\n",
              "       0.09035   , 0.09863333, 0.10441667, 0.09751667, 0.09915   ])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "T6BSWTAVJPvh",
        "colab_type": "text"
      },
      "source": [
        "Let's now calculate the Log of the Gaussian likelihood. We can use the following equation:\n",
        "\n",
        "\\begin{equation}\n",
        "\\large\n",
        "\\ln\\mathcal N(\\mathbf x;\\mu,\\sigma)=\\sum_{i=1}^D\\ln\\sigma_i-\\frac 1 2 \\sum_{i=1}^D \\left(\\frac {x_i-\\mu_i}{\\sigma_i}\\right)^2\n",
        "\\end{equation}"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sr08kJ57DXky",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def log_gaussian(X, mu, sigma):\n",
        "    return -(np.sum(np.log(sigma)) + 0.5 * np.sum(((X - mu) / sigma) ** 2)).reshape(-1, 1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HwDDtJXxK6ru",
        "colab_type": "text"
      },
      "source": [
        "The following function allows us to train our model. This process is all about finding the class-level means $\\mu$ and standard deviations $\\sigma$. Once we know these parameters, then we can easily calculate the predictions by replacing their values in the Gaussian density function."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hVbWcvcGy_ho",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def train_naive_bayes():\n",
        "    means = np.zeros((num_classes, X_train.shape[1]), dtype=np.float64)\n",
        "    stdevs = np.zeros((num_classes, X_train.shape[1]), dtype=np.float64)\n",
        "    \n",
        "    for k, y_k in enumerate(np.unique(y_train)):\n",
        "        indices = np.where(y_k == y_train)\n",
        "        means[k] = np.mean(X_train[indices], axis=0)\n",
        "        stdevs[k] = np.std(X_train[indices], axis=0)\n",
        "        \n",
        "    return means, stdevs"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nSHLe1h2Lcfq",
        "colab_type": "text"
      },
      "source": [
        "Now let's train our model! At this point, we will also display the class-level mean and covariance for the first category of our dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZUvfX_fJ5mqp",
        "colab_type": "code",
        "outputId": "112bdfaf-f83e-4314-8583-405a7dbfcb7f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 680
        }
      },
      "source": [
        "means, stdevs = train_naive_bayes()\n",
        "print('Means for class 0:\\n', means[0], '\\n\\nStandard deviations for class 0\\n', stdevs[0])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Means for class 0:\n",
            " [0.30678743 0.39533906 0.49698065 0.47942126 0.23345504 0.16344804\n",
            " 0.09279153 0.01742686 0.01565332 0.01517458 0.01731597 0.17423633\n",
            " 0.44260052 0.43626217 0.43094108 0.29247752 0.06933765 0.05582165\n",
            " 0.04511124 0.02725237 0.06435403 0.12624306 0.15242024 0.3932381\n",
            " 0.33039128 0.15454434 0.12076002 0.07976528 0.09248541 0.24739966\n",
            " 0.29079921 0.26497931 0.36151173 0.36524211 0.31475686 0.3574081\n",
            " 0.17230083 0.19654137 0.36610455 0.38814007 0.32552076 0.35568275\n",
            " 0.32603948 0.18848521 0.22032257 0.17830103 0.092407   0.14042486\n",
            " 0.32399185 0.33429498 0.39768063 0.37237062 0.24249547 0.23246207\n",
            " 0.18654091 0.0988651  0.13513616 0.15262434 0.13333062 0.26915728\n",
            " 0.10747293 0.10649556 0.29421138 0.40128702 0.39745421 0.41461391\n",
            " 0.38427027 0.18927837 0.19220744 0.12930194 0.06576277 0.09647302\n",
            " 0.27370977 0.08789382 0.09359048 0.11626488 0.10110692 0.23180832\n",
            " 0.27630185 0.25474979 0.37876405 0.39644995 0.34596988 0.34044592\n",
            " 0.42185483 0.37627579 0.37949988 0.26628861 0.08103729 0.06562463\n",
            " 0.05504187 0.04430536 0.13012485 0.23477544 0.26115985 0.40948865\n",
            " 0.28233761 0.35860327 0.48115529 0.48099913 0.33998553 0.22773549\n",
            " 0.10881633 0.00711425 0.00496192 0.00806203 0.01549253 0.15777465] \n",
            "\n",
            "Standard deviations for class 0\n",
            " [0.15528214 0.13907786 0.07806083 0.10262917 0.17125935 0.13673659\n",
            " 0.09997971 0.04588158 0.04576941 0.0486846  0.05062703 0.13881413\n",
            " 0.07057316 0.08547206 0.09635951 0.17286217 0.09686256 0.07293858\n",
            " 0.06116716 0.04474594 0.08637365 0.12940148 0.14241601 0.11331706\n",
            " 0.08383711 0.12410426 0.12281056 0.10394103 0.10687686 0.1123704\n",
            " 0.0974079  0.12141028 0.07698211 0.07799513 0.11595225 0.06842645\n",
            " 0.12034932 0.1324934  0.10633574 0.0925468  0.12955381 0.08791516\n",
            " 0.09726449 0.13325494 0.13515623 0.1305467  0.09427651 0.11038917\n",
            " 0.12147878 0.12776927 0.09511013 0.11812663 0.14998211 0.13703613\n",
            " 0.1309343  0.1104304  0.12818632 0.13377738 0.12666196 0.13428387\n",
            " 0.10187655 0.10869919 0.14785681 0.08905034 0.09840079 0.0687211\n",
            " 0.08940851 0.14359557 0.13969207 0.12103721 0.07948856 0.09529809\n",
            " 0.12384352 0.1074754  0.11095951 0.11749306 0.10375126 0.11423364\n",
            " 0.11047158 0.14045158 0.10451257 0.08677471 0.11171208 0.09933593\n",
            " 0.06044544 0.1084184  0.1008463  0.14537294 0.09439435 0.07531606\n",
            " 0.06742613 0.06649292 0.12993194 0.14986582 0.14983616 0.0732667\n",
            " 0.14264207 0.13925709 0.06555377 0.06561917 0.14893103 0.14506146\n",
            " 0.10164724 0.02301921 0.01833006 0.02644523 0.03746856 0.12064316]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DsBYCIPhLz-v",
        "colab_type": "text"
      },
      "source": [
        "The following function allows us to make predictions based on our parameters just found."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RUqS1rCS9mpk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def predict(x):\n",
        "    return np.argmax([np.log(prob_y[k]) + log_gaussian(x.reshape(1, -1), \n",
        "                      means[k], stdevs[k]) for k in range(num_classes)])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1AImUel4L9lr",
        "colab_type": "text"
      },
      "source": [
        "Now let's find the accuracy of our model by predicting the __y__ value for each image in our dataset. This will be 91.5% which is just fine. That means that even with the independence assumption we can get good prediction results. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "utbHFN8DC2hS",
        "colab_type": "code",
        "outputId": "8165d1e4-fe40-4a1e-c481-dfa615625fa6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_pred = np.zeros((X_test.shape[0]))\n",
        "\n",
        "for n in range(X_test.shape[0]):\n",
        "    y_pred[n] = predict(X_test[n])\n",
        "\n",
        "accuracy_score(y_test, y_pred)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9151"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xNHcTB70MMPw",
        "colab_type": "text"
      },
      "source": [
        "Let's now display some items from the testing dataset with their corresponding predictions."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jYq-oFNMNi6i",
        "colab_type": "code",
        "outputId": "873fa634-ec6c-43aa-d37f-7577e5c819f5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 576
        }
      },
      "source": [
        "_, ax = plt.subplots(6, 8, figsize=(10, 10))\n",
        "index = 0\n",
        "\n",
        "for i in range(6):\n",
        "    for j in range(8):\n",
        "        cell = ax[i][j]\n",
        "        cell.set_title('Pred: %s' % predict(X_test[index]))\n",
        "        cell.set_xticks([])\n",
        "        cell.set_yticks([])\n",
        "        cell.grid(False)\n",
        "        cell.imshow(X_raw_test[index], cmap='gray')\n",
        "        index += 1\n",
        "        \n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkMAAAIvCAYAAAB3D9+6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
            "text/plain": [
              "<Figure size 720x720 with 48 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "26STbyS4Mrmb",
        "colab_type": "text"
      },
      "source": [
        "Moving forward, we also compare our result with sklearn's. In this case, we obtain exactly the same accuracy value of 91.5%"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R26XT9JW--li",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.naive_bayes import GaussianNB"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tM5YW1uN_Ng2",
        "colab_type": "code",
        "outputId": "d67acfdd-af1c-4240-d6a5-27fd588d26c8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "model = GaussianNB()\n",
        "model.fit(X_train, y_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GaussianNB(priors=None, var_smoothing=1e-09)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CH1TFtc-_XE3",
        "colab_type": "code",
        "outputId": "2e43c381-9187-4a75-ddb4-635e3198f65a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_pred = model.predict(X_test)\n",
        "accuracy_score(y_test, y_pred)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9151"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    }
  ]
}