Compute Q-Value Estimation for RL in MuJoCo environment.

Implementing reinforcement learning for an agent and crawler. Both learn by exploring different paths to reach the goal state. Ultimately, both are able to find optimal paths. We can try both manual paths or simulate a number of episodes.

This project implements a Deep Q-Network (DQN) reinforcement learning agent to solve the LunarLander-v3 environment provided by the Gymnasium library. The agent learns to land a lunar module smoothly between two flags by maximizing cumulative rewards using trial-and-error interaction.