This is a simple Machine Learning project on recognition of hand written digits. A feedforward neural network model is trained on MNIST dataset to classify handwritten digits 0-9. The main focus of the project was to create a web application to deploy the project. A python web application was built using Flask. The canvas API of javascript is used to draw digits on the web page. The digit is saved as a image and used as input to the model and the digit is predicted. The predicted number is displayed on the screen.
