This project is about applying common Machine Learning models to a Hand digit classification problem with/without dimension reduction using PCA

Goal of the project

• Applying and comparing the results of K-nearest-neighbors, Classification-Tree, Random Forest, Support vector machine and Neural Network with a Hand Digit classification problem.

• All of the above methods will be applied to original data and dimention-reducted data gotton by applying Principal Component Analysis (PCA).

Link to the presentation and R code files

Here is the link to presentation file Here is the link to the R code files

Convolution Neural Networks with the MINIST data

After finishing the Project that applied basic Machine Learning to the data. I had move to learn Convolution Neural Networks (CNN) and Bayesian Convolution Neural Networks (BCNN) from the book Probabilistic Deep Learning. I have learned that applying the CNN can help getting 97% accuracy on the test set. Here is the link to the jupyter notebook