Basics of Neural Networks in Go
I've decided to start this project as a way to put together my enthusiasm for Go together an opportunity to learn a bit.
This is a very simplistic Neuron. It accepts a 2-dimensional array, n features and m examples.
The code should be pretty easy to read in itself.
To try it out, just initialize the neuron with the input and output of your liking.
At some point I might move the input and output to be set from the command line but for now just change it in the main function. the same goes for the number of iterations, the number of features and the learning rate.
The idea is to keep adding features so that at some point you can actually do some interesting things, but this is by no means focused on speed or scalability. The main focus is on code readiability.
Things in my TODO list:
- Allowing Neurons to be stacked in layers
- Add Optimization Algorithms (Adams)
- Add Activation Functions (ReLU, Sigmoid, Tanh, Softmax)
I'm also open to suggestions.
DISCLAIMER: This was created for educational purposes. Feel free to use this code but use it at your own discretion.
This was originally inspired by this blog post by @iamtrask.