Research
Exploring intelligent systems.
I'm currently pursuing my Master's in Computer Science at the College of Staten Island, with a focus on artificial intelligence and machine learning.
My current research focuses on reinforcement learning and multi-agent decision-making, particularly how different learning approaches perform in strategic environments.
Current Research
Master's Thesis
Evaluating Actor-Critic Reinforcement Learning for Multi-Agent Decision-Making in So Long Sucker
College of Staten Island
Multi-agent reinforcement learning provides a framework for studying decision-making among interacting agents in competitive and cooperative environments. So Long Sucker is a four-player strategic game involving competition, temporary coalitions, and changing player incentives.
Previous research introduced the game as a multi-agent reinforcement learning benchmark and evaluated value-based methods including Deep Q-Networks (DQN), Double DQN (DDQN), and Dueling DQN.
My research extends this work by evaluating actor-critic reinforcement learning methods and comparing their performance with these previously tested value-based approaches.
Approach
01
Reproduce
Replicate the existing value-based experiments to evaluate reproducibility and establish experimental baselines.
02
Extend
Implement selected actor-critic reinforcement learning methods within the existing So Long Sucker environment.
03
Compare
Evaluate value-based and actor-critic approaches under comparable experimental conditions and analyze their relative performance.