This repository contains my submission for the second project of the Udacity Artificial Intelligence Nanodegree program. In this project we game playing agent to play the Isolation game as knight (able to make the same moves as a knight from the game Chess). In order to create the agent I implemented the Minimax and AlphaBeta search algorithms. These adversarial search algorithms search future player moves to find the best move according to an evaluation heuristic that maximizes the player's positional value and minimizes the opponents positional value. I also evaluated several heuristics and analyzed their performance in the heuristic_analysis.pdf file. This project submission also required a summary of an artificial intelligence research paper related to game playing. I chose to create a two page summary of the "Mastering the game of Go with deep neural networks and tree search" paper by DeepMind1 (please seeresearch_review.pdf). The objective of the paper is to summarize the key innovations made by the DeepMind team which allowed their AlphaGo machine to become the best Go player in the world.
Created by Udacity Staff.
This project requires Python 3. Project used a pre-packaged Anaconda3 environment provided by the Udacity staff for the artificial intelligence nanodegree program.
game_agent.py- Contains the code for the Isolation playing agent.agent_test.py- Provided by the Udacity staff for functional testing (runpython agent_test.py).tournament.py- Provided by the Udacity staff for performance evaluation of the Isolation playing agent (runpython tournament.py).
[1] Mastering the game of Go with deep neural networks and tree search, by David Silver et als @ https://storage.googleapis.com/deepmind-media/alphago/AlphaGoNaturePaper.pdf