Our work aims to conduct a comparative analysis of reinforcement learning algorithms using the highway-environment developed by OpenAI Gymnasium. Specifically, we focus on three main methods: Deep Q-learning, Proximal Policy Optimization (PPO), and Actor-Critic. While previous research exists on the highway-environment, there is limited implementation of these specific algorithms. We believe it would be valuable to implement these algorithms from scratch and evaluate their performance on this environment, comparing them.
The primary objective of this project is to compare the effectiveness and performance of Deep Q-learning, PPO, and Actor-Critic algorithms in navigating the highway-environment. By implementing these algorithms from scratch and conducting experiments, we seek to gain insights into their strengths, weaknesses, and suitability for autonomous driving scenarios.
- Deep Q-learning, Actor-Critic Methods and Proximal Policy Optimization are compared
- Implementation of algorithms from scracth
- Evaluation of algorithms' performance on the highway-environment.
- Generation of insights into algorithm behavior and effectiveness in autonomous driving scenarios.