This project takes a picture of a digit that has a size of 28 by 28 and will convert that to number from 0 to 9 that represents the value of the picture.

This project was made to submit for the https://www.kaggle.com/c/digit-recognizer

This project currently has an accuracy of 0.97657 and is currently around ~300th place in the kaggle competition.

This works by using a neural network with 3 layers. There is the input layer, the output layer and the hidden layer. The input Layer has a size of 28*28 which takes in the value of every single value of each pixel in a black and white mode. There are 150 hidden layer nodes. The output layer has a size of 10. There are 10 output layers because of the 10 different possible outputs you can get. Each neuron is responsible for a number and the one that got fired is the result