The first mini-project of COMP 551 - Applied Machine Learning.
-
The project description is available here: http://www.reirab.com/Teaching/AML20/P1.pdf
-
Tests of logistic regression and naive Bayes for the majority of datasets: http://ai.stanford.edu/~ang/papers/nips01-discriminativegenerative.pdf
-
Paper related to the 3rd part (Ng AY, Jordan MI. On discriminative vs. generative classifiers: A comparison of logistic regression and naive bayes. In Advances in neural information processing systems 2002 (pp. 841-848)): https://papers.nips.cc/paper/2020-on-discriminative-vs-generative-classifiers-a-comparison-of-logistic-regression-and-naive-bayes.pdf
Below are the files that cover the whole project. The other files are supplimentary and served as a playground for testing.
- COMP 551 - Project 1 - Adult.ipynb
- COMP 551 - Project 1 - Chess (King-Rook vs. King-Pawn).ipynb
- COMP 551 - Project 1 - Ionosphere.ipynb
- COMP 551 - Project 1 - Tic-Tac-Toe Endgame.ipynb
- writeup.pdf
- Completeness: 20/20
- Corrnectness: 32/40
- (-2) The experiments on different training sizes is not correct: you should fixed the size of the test set (e.g. 20% of all data), then change the size of the training set (e.g. 100%/90%/80%/70% of the complete 80% training set)
- Writing: 20/25
- exceed the page requirements (5 v.s. 11) (-5) (p.s. you should put all extra things in an appendix. If you are submitting to a conference, exceeding required pages will be rejected directly....)
- Originality: 8/15
- detailed data analysis on the four datasets (+8)
Total: 83%
The repositary was private for the duration of developement and was made public on Apr 16, 2020.