AdalbertoCq / Reinforcement-Learning-Agents

Deep Deterministic Policy Gradient, Double Deep Q Learning and Deep Q Learning Agents.

Reinforcement Learning Agents

• Deep Q-Learning:

  • Based on V.Mnih et al. "Playing Atari with Deep Reinforcement Learning", 2013
  • Deep Q-Learning implementation.
  • Implementations of neural network action-value approximator in TensorFlow.
  • Implemented experience replay memory and fixed Q targets.
  • CartPole v0 OpenAI gym Q Rewards & Q value NN loss.

• Double Deep Q-Learning:

  • Based on H.Hasselt et al. "Deep Reinforcement Learning with Double Q-learning", 2015
  • Double Deep Q-Learning implementation.
  • Implemented experience replay memory and fixed Q targets.
  • Implemented two action-value neural network approximators, for action decision and fixed target.
  • CartPole v0 OpenAI gym Q Rewards & Q value NN loss.

• Deep Deterministic Policy Gradient:

  • Based on T.Lillicrap et al. "Continuous control with deep reinforcement learning", 2016
  • Deep Deterministic Policy Gradient implementation.
  • Implemented action repeat, experience replay memory and fixed targets for Actor/Critic Networks with soft update.
  • MountainCarContinuous-v0 solved after 70 episodes.
  • MountainCarContinuous-v0 Critic Loss and Rewards.