/Lunar_RL

Comparison of performance of different reinforcement learning algorithms applied to the lunar lander problem

Primary LanguagePython

AE504 Final Project

Authors: Gokul Puthumanaillam, Manav Vora

This is a code for comparing different algorithms to solve the lunar lander problem. The algorithms simulated are:

  1. Monte Carlo
  2. Q-Learning
  3. DQN
  4. SARSA

Outputs:

Simulation Final Trajectory
randomRandom policy random_plot
monte_carloMonte Carlo monte_carlo
q_learningQLearning q_learning
sarsaSARSA sarsa
dqnDQN dqn