Sine-based_neural_network_MNIST_Processing

This is a function oriented sine-based neural network (SNN) written in processing without using libraries. It is structured in a easy to read fashion hoping to help beginners better understand how a neural network functions as a few arrays of weights. I wrote this when I was trying to learn to code my own neural network and found it hard to track all the functions and variables in Object Oriented Programing. So I wrote this one in functions. And I structured the code following the shape of this SNN so it is easy on brains. The SNN is invented by Tekin Evrim Özmermer and described as formular 2 in https://www.researchgate.net/publication/343505521_Sinusoidal_Neural_Networks_Towards_ANN_that_Learns_Faster You can also find his github entry following https://github.com/evrimozmermer/sinusoidal_neural_networks_experiments_4_1 The reduced MNIST and method to load data are from Charles Fried's tutorial "Let's code a Neural Network from scratch" at https://medium.com/typeme/lets-code-a-neural-network-from-scratch-part-1-24f0a30d7d62