ML assignment 3: Made multiple CNN based models, and optimised parameters to get the best accuracy on validation and test data

This repo only contains the code and the report. The report contains the details of my experiments along with the graphs too. This project conatined various parts:

1. Classification using Kmeans clustering
2. Classification using one vs one classifier of SVM after reducing the dimensions of the data using PCA
3. Using NN
4. Using CNN and its various combinations to get best possible output.

The code for all this is submitted in form of Jupyter notebooks and python files.