Wide and Deep Reinforcement Learning Extended for Grid-Based Action Games.
Juan M. MontoyaChristoph DoellChristian BorgeltPublished in: ICAART (Revised Selected Papers) (2019)
Keyphrases
- reinforcement learning
- action sets
- learning agents
- action selection
- extensive form games
- action space
- wide range
- reinforcement learning agents
- state space
- markov decision processes
- partially observable domains
- optimal policy
- reward shaping
- state action
- function approximation
- game theory
- neural network
- temporal difference
- game theoretic
- dynamic programming
- transition model
- computer games
- fitted q iteration
- model free
- multiagent learning
- monte carlo tree search
- game playing
- game play