In this assignment, you will build a complete neural network using Numpy. You will implement all the steps required to build a network - feedforward, loss computation, backpropagation, weight updates etc. You will use the MNIST dataset to train your model to classify handwritten digits between 0-9.
The assignment is divided into the following sections:
- Data preparation
- Feedforward
- Loss computation
- Backpropagation
- Parameter updates
- Model training and predictions