Purple rAIn

The simplest neural network in existence (2 lines of code with no 3-party dependency)

What it does? Classifies if the given RGB input resolves to purple or orange (approximately)

Run:

> python .\purple_rain.py "150,100,150"
> purple

Graph of the neural network with 2 layers:

---

but wait...

yes, they are.

The neural network encodes multiple choices into a single matrix (your model) whereby each node is an if condition and each weight defines the likelihood to activate the if condition (neuron) for any given input.

PS: of course this is a very poor AI with very low accuracy. It can only categorize 2 colors and the weights were defined at my own taste (they barely matter).

DO NOT USE IN PRODUCTION