Q$.3$ You are tasked with building a feedforward neural network from scratch to classify images

of handwritten digits from the MNIST dataset. The network should have one hidden layer and use

the sigmoid activation function for the hidden layer and the softmax activation function for the

output layer. Here are the specifications:

$[6$ Marks$]$

Input layer: $784$ nodes $($corresponding to the $2828$ pixel images$)$

Hidden layer: $128$ nodes with sigmoid activation function

Output layer: $10$ nodes $($for digits $0$ to $9)$ with softmax activation function

Loss function: Categorical Cross$-$Entropy