A simple set of files to reproduce the concept of Perceptron, solving the position of points relative to a given linear function, and combining 3 elements to solve the non separable XOR problem
Script runs over main.py - function can be changed inside Utility.py
Perceptron is trained one datapoint per frame, clicking the mouse left button triggers a training on the whole batch for one frame
PS: The whole code is not well commented, surely not written the best way possible, probably not the fastest (performance wise) possible. Just my attempt to understand neurons and weights paradigm :)