A repository for reinforcement learning projects, specifically q-learning.
The frozen lake example in this repository is a simple implementation of the q learning algorithm. The purpose of the frozen lake game is for a reinforcement learning agent to navigate it's way across a grid of safe and unsafe tiles. If the agent is successful it is rewarded, otherwise it is not. The goal of the agent is to select the set of actions that maximizes the reward it obtains thus 'learning' which actions are best based on the state the agent/environment is in.
A detailed description of the q learning algorithm is prorvided in the pdf slide show.
This project was part of a Data Science and Analytics course at the University of Toronto.