Reinforcement Learning Baseline

We tend to implement stable versions of popular deep reinforcement learning algorithms and test them in various problems.

Algorithms:

Cross-Entropy Method (CEM)

• paper: The Differentiable Cross-Entropy Method
• implementation: Ivan Pavelyev, Anton Plaksin

Deep Q-Network (DQN)

• paper: Playing Atari with Deep Reinforcement Learning

Double Deep Q-Network (DDQN)

• paper: Deep Reinforcement Learning with Double Q-learning

Deep Deterministic Policy Gradient (DDPG)

• paper: Continuous Control with Deep Reinforcement Learning

Normalized Adavtage Functions (NAF)

• paper: Continuous Deep Q-Learning with Model-based Acceleration
• implementation: Stepan Martyanov

Asynchronous Advantage Actor-Critic (A3C)

• paper: Asynchronous Methods for Deep Reinforcement Learning
• implementation: Alexander Chaikov

Continuous Value Iteration (CVI)

• paper: Value Iteration in Continuous Actions, States and Time
• implementation: Stepan Martyanov

Proximal Policy Optimization (PPO)

• paper: Proximal Policy Optimization Algorithmse
• implementation: Viktor Sergeev

Soft Actor-Critic (SAC)

• papers: Learning to Walk via Deep Reinforcement Learning
• implementation: Anton Plaksin