This project is focussed at classification of 0-9 handwritten digits using MNIST dataset, using Artificial Neural Network (ANN).
- The training dataset contains 60,000 handwritten digit images, each of 28x28 pixels.
- File : train-images.idx3-ubyte
- The labels of these images are stored in seperate file
- File : train-labels.idx1-ubyte
- The test dataset contains 10,000 handwritten digit images, each of 28x28 pixels.
- File : t10k-images.idx3-ubyte
- The labels of these test images are stored in seperate file
- File : t10k-labels.idx1-ubyte