knnMNIST

This project is part of the Statistical Learning Theory course I attended at the TU/e Eindhoven University of Technology. Implementation of the kNN (k-Nearest Neighbors) algorithm for classification of handwritten digits on the MNIST dataset. The report describe and evaluate tour implementation considering problems such as its performance, the suitable number of neighbors, and the best distance_metric. We also applied PCA (Principal Component Analysis) and some image-blurring techniques on the image to improve classification accuracy.