Implemented these two Machine Learning algorithms while studying CS 189 @ UC Berkeley.
Project Outline:
classifier.py
- machine learning code of LDA and QDAtrain.py
- sample training code using LDA or QDA with MNIST or SPAMdata/spam.mat
- matlab file containing spam training data- each row represents an email labeled as either "spam" or "ham"
- each col represents frequency of a word (actual data is l1-normalized)
data/spam.obj
- python pickle file containing dictionary of wordsdata/mnist.mat
- matlab file containing mnist digit training data- each row represents a 28-by-28 flatten greyscale image
- each col represents greyscale from 0 to 255 (actual data is l2-normalized)
The data files are not included in this github repo, they can be downloaded here.