ii0 / reinforcementlearning

A simple Reinforcement Learning project written in Swift

Reinforcement Learning

This is a simple Reinforcement Learning project written in Swift.

Basically, it navigates around a grid to find the goal (without knowing it before hand). Then it constructs an optimal path to the goal using model free reinforcement learning.

Here's a visual representation:

1. Green color represents Q-value
2. Red color represents failure node
3. Goal is to reach top-right corner

Please see the following files for more information:

1. QLearning.swift
2. AISolver.swift