Udacity-AIND-Adversarial-Search | Adversarial Game Playing Agent for Knights Isolation

The easiest way to complete the project is to use the Udacity Workspace in your classroom. The environment has already been configured with the required files and libraries to support the project. If you decide to use the Workspace, then you do NOT need to perform any of the setup steps for this project. Skip to the section with instructions for completing the project.
If you would prefer to complete the exercise in your own local environment, then follow the steps below:.
Open your terminal and activate the aind conda environment (OS X or Unix/Linux users use the command shown; Windows users only run activate aind)
Download a copy of the project files from GitHub and navigate to the project folder. (Note: if you've previously downloaded the repository for another project then you can skip the clone command. However, you should run git pull to receive any project updates since you cloned the repository.)
Create a performance baseline using run_search.py to evaluate the effectiveness of a baseline agent (e.g., an agent using your minimax or alpha-beta search code from the classroom)
Use run_search.py to evaluate the effectiveness of your agent using your own custom search techniques
You must decide whether to test with or without "fair" matches enabled--justify your choice in your report
If the results are very close, try increasing the number of matches (e.g., >100) to increase your confidence in the results
Experiment with adding more search time--does adding time confer any advantage to your agent?
Augment the code to count the nubmer of nodes your agent searches--does your agent have an advantage compared to the baseline search algorithm you chose?
You MAY implement advanced techniques from the reading list at the end of the lesson (like Monte Carlo Tree Search, principle variation search, etc.), but your agent is being evaluated for performance rather than correctness. It's possible to pass the project requirements without using these advanced techniques, so project reviewers may encourage you to implement a simpler solution if you are struggling with correct implementation. (That's good general advice: do the simplest thing first, and only add complexity when you must.)