CapstoneProject — NEAT Tetris Agents (Unity / C#)

This project contains multiple AI agents built to play Tetris in Unity (C#). The agents use NEAT (NeuroEvolution of Augmenting Topologies): a neural network evolved via a genetic algorithm.

What’s in this repo

There are 3 versions of the agent (implemented on separate branches):

ANGELINA (branch: Version-2)

• Evaluates gameplay move-by-move (not only at the end of the game)
• For each new piece, evaluates candidate placements to choose a target board state
• Uses A* search to route the piece to the chosen placement

BRAD V1 (branch: Version-1.1)

• Evaluates the gameboard after the game ends
• Each agent plays one full game and is scored by that performance

BRAD V2 (branch: Version-1.2)

• Evaluates the gameboard after the game ends
• Each agent plays three games and is scored by the average performance

The master branch contains the project baseline; the main agent implementations live in the branches above.

Tech stack

• Unity (Tetris environment)
• C# (NEAT implementation + genetic algorithm + agent logic)

How to run

1. Install Unity (use any version that can open the project in the selected branch).
2. Clone the repo:

git clone https://github.com/keniel123/CapstoneProject.git
cd CapstoneProject

3. Check out the version you want:

# Angelina
git checkout Version-2

# Brad V1
git checkout Version-1.1

# Brad V2
git checkout Version-1.2

4. Open the project folder in Unity.
5. Press Play to run the environment (exact scene / entrypoint depends on the selected branch).

Collaborators

• Keniel Peart
• Kimberly Soares
• Rajay Bitter
• Shanice Bryan

License

See LICENSE.