Code for ML students.
author = Johannes Bjerva and Malvina Nissim
version = 0.3
license = GPL v3
maintainer = Mike Zhang
status = Testing
First, we are going to install a package manager. This makes sure that all your packages are up-to-date:
https://docs.conda.io/en/latest/miniconda.html
Please note: pick the Python 3.8 version, also check whether your system is 32-bit or 64-bit.
After installing miniconda, we are going to install some packages:
(1) Go to your terminal (Anaconda prompt (miniconda3) for Windows. Windows users go to (3)).
(2) Type (or copy-paste):
conda activate
(3) Type (or copy-paste):
conda install matplotlib numpy pandas scikit-learn
If you see the Proceed ([y]/n)? prompt: press y to continue.
https://github.com/jjzha/foml/archive/master.zip
(1) Do only once:
git clone https://github.com/jjzha/foml.git
(2) To get most recent version:
git pull
The difference between static and dynamic is that every time we update some code, you have to download the new code manually. By using git pull you are able to pull the code from the repository.
(1) go to your foml directory (by using cd command). For example:
cd ~/Downloads/foml-master
It might be that your foml-master directory is not in the Downloads folder.
To check where you are currently, type in the terminal/command prompt: pwd (for Linux/MacOS users) or echo %cd% (for Windows users).
Then try to navigate to your folder where the foml directory is with cd (for Linux/MacOS/Windows).
(2) to run an experiment, try:
python run_experiment.py --csv data/hyperp_subset.csv --algorithms nb --features text-cat --nwords 1 --split 70 30
where the following arguments mean:
--csv <string>is the path to yourcsvfile.--algorithms <string>is the abbrevation of your algorithm (e.g. Naive Bayes = nb).--features <string>is the columns you want to use as features.--nwords <int>is the number of words you want the system to look at.--split <int, int>is the split you want to use for training and testing, in the example it is 70% training and 30% test, hence70 30. By default, it will use the 70% training set for training and development. If the--testargument is given, it will show the result on the test set.
conda install <package_name>
Replace <package_name> with your preferred package.
conda update conda
Having a conda environment on your pc is very convenient. You are able to run your scripts in this environment without having python locally.
With a focus on NLP:
- Christopher D. Manning and Hinrich Schütze, Foundations of Statistical Natural Language Processing, MIT Press. Cambridge, MA. 1999. http://nlp.stanford.edu/fsnlp/
- Christopher D. Manning, Prabhakar Raghavan and Hinrich Schütze, Introduction to Information Retrieval, Cambridge University Press. 2008. http://nlp.stanford.edu/IR-book/
- James Pustejovsky and Amber Stubbs, Natural Language Annotation for Machine Learning, O’Reilly. 2012.
- Steven Bird, Ewan Klein, and Edward Loper, Natural Language Processing with Python, O’Reilly. 2009. http://www.nltk.org
- Hal Daumé III. A course in Machine Learning. http://ciml.info (incomplete manuscript available online – some parts available for free.)
More generally on machine learning:
- Tom Mitchell, Machine Learning, McGraw Hill. 1997.
- Ian H. Witten, Eibe Frank, Mark A. Hall, Data Mining: Practical Machine Learning Tools and Techniques, The Morgan Kaufmann Series in Data Management Systems. 2011.
- Yaser S. Abu-Mostafa, Malik Magdon-Ismail, Hsuan-Tien Lin, Learning from Data, AMLBook. 2012.
- Peter Flach, Machine Learning: The Art and Science of Algorithms that Make Sense of Data, Cambridge University Press. 2012.
More specific to Scikit learn (and ML with Python):
- Luis Pedro Coehlo and Willi Richert, Building Machine Learning Systems with Python, PACKT Publishing. 2013.
- Raúl Garreta and Guillermo Moncecchi, Learning scikit-learn: Machine Learning in Python, PACKT Publishing. 2013.
More into background of neural nets:
If you have any questions, please ask during the lab or contact me via e-mail: mikz@itu.dk