AI Plays (Classical Methods)

This repository is a collection of various Classical AI and Reinforcement Learning algorithms for building agents that can play simple games or solve problems. This repo is a precursor to a future collection of Deep Reinforcement Learning Algorithms.

Contents

1. Implementation of Two Player TicTacToe Game
2. Solving N-Queens (a Constraint Satisfaction Problem) with Backtracking
3. Solving Mazes with Various Search Algorithms

TODO

A checklist of things to do in this repository.

• Code a Normal 2 Player TicTacToe Terminal Game
• Solving a Constraint Satisfaction Problem (NQueens) using Backtracking
• Maze Solving with DFS and BFS
• Maze Solving with A* Search
• Monte Carlo Methods for Blackjack
• Create Gym Environment for TicTacToe
• MinMax Algorithm and Self-Play
• NeuroEvolution
• Classical Q Learning