yunjiaxi/simple-rl-practice

Some basic RL algorithm include DP, MC, TD, SARSA, Q-Learning, DQN, A3C

Simple-RL-Practice

Some basic RL algorithms including DP, MC, TD, SARSA, Q-Learning, DQN, A3C

Content

• DP (Policy Evaluation & Policy Iteration & Value Iteration)
• MC & TD (First-visit & Every-visit MC, TD(0))
• Model Free Control (SARSA & Sarsa(lambda), Q-Learning)
• DQN (basic DQN, double DQN, dueling DQN, PER DQN)
• A3C (RMSprop, Adam, SharedRMSprop, SharedAdam)

Dependencies

• python 3.6
• numpy
• pytorch 1.1.0
• gym
• matplotlib