Neural Network

This is my attempt to code a neural network from scratch after studying how deep learning works

Activation functions

• Linear (Identity)
• Sigmoid (Logistic)
• ReLU (Rectified linear unit)
• TanH

Loss functions

• MSE (Mean squared error)
• Binary cross entropy

Example: XOR

To test my module, I tried to train a network predicting the XOR gate output

Cost over epochs

Decision boundary output