Expanding Activation Functions Module

[jrego43]

Activation functions are the key to representing complex nonlinear functions with neural networks. As such, it is important that we implement a wide variety of them to ensure a flexible framework. You can see many activation functions here that should be considered for implementation. As of now, we have ReLU, Softmax, and Tanh completed, so you can use their source files for reference.

[jrego43]

I'll be working on the sigmoid activation function starting tomorrow.

⬇️

[Pochetes]

update from signal: @jrego43 will be working on #3 and I'll be doing this one. slight confusion there