Implementation of Naive Bayes, Gaussian Naive Bayes, and 5-fold cross-validation Nearest neighbor with pure python.
The data that is used is MNIST dataset: http://yann.lecun.com/exdb/mnist/
main.py file have three functions:
Problem1: implements Naive Bayes on mnist for 10 classes (0-9), after making data binary. (creating a threshhold between 0 - 255)
Problem2: implement Gaussian Naive Bayes onmnist for two classes (5 or not 5), on actual data (from 0 - 255).
Problem3: Creates 5-fold Cross validation Nearest neighbour classifier on 3 classes (1,2,7) with different K and chooses best K for test set.