Some types of Neural networks:
Activation function
the purpose of the activation function is to introduce non-linearity into the output of a neuron, enabling the network to learn complex patterns, and in a Deep Learning setting, this means the model can pick up on patters without manual Feature Engineering.