neuralnet-with-backpropagation

This neural network has an input layer, one hidden layer, and one output layer. The input layer - 784 nodes, the hidden layer - 128 nodes and the output layer - 10 nodes.

The sigmoid function is used as activation function.

Using MNIST Dataset

Implementation:

1. Forward propagation (forward pass)
2. Backward propagation (backward pass)
3. Updating the gradient

Using only the following functions:

• torch.add • torch.mul • torch.transpose • torch.mm

Run Using Google Colab