Using Generative Adversarial Nets on Atari Games for Feature Extraction in Deep Reinforcement Learning.
Ayberk AydinElif SürerPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- linear value function approximation
- feature extraction
- reinforcement learning problems
- multi agent
- reinforcement learning algorithms
- generative model
- function approximation
- game theory
- minimax search
- state space
- optimal policy
- markov decision problems
- nash equilibria
- learning algorithm
- model free
- temporal difference
- stochastic games
- temporal difference learning
- markov decision processes
- support vector machine svm
- machine learning
- game playing
- learning capabilities
- control problems
- policy iteration
- reinforcement learning methods
- transfer learning
- markov games
- game theoretic
- fixed point