/Handwritten-Digit-Classifier

Classification on MNIST dataset

Primary LanguageJupyter Notebook

Handwritten-Digit-Classifier

This is an example of classification of handwritten digits 0 to 9 on MNIST dataset.

To run this jupyter notebook, download MNIST dataset from Kaggle.Use below link to download the dataset. https://www.kaggle.com/vikramtiwari/mnist-numpy

Extract the dataset in the same folder as this notebook.

This deep neural network can give 99.13% accuracy on train images and 99.12% accuracy on test images.

After adding convolution and pooling layer in the neural network, model gives 99.84% accuracy on train and test images.