Deep Pepper: Expert Iteration based Chess agent in the Reinforcement Learning Setting.
Sai Krishna G. V.Kyle GoyetteAhmad ChamseddineBreandan ConsidinePublished in: CoRR (2018)
Keyphrases
- reinforcement learning
- multi agent
- action selection
- learning agent
- state abstraction
- reward function
- evaluation function
- partially observable
- multi agent systems
- intelligent agents
- human players
- multiagent systems
- state space
- temporal difference
- function approximation
- state action
- autonomous learning
- agent receives
- inverse reinforcement learning
- decision making
- multi agent environments
- artificial intelligence
- learning agents
- general game playing
- agent technology
- markov decision process
- single agent
- learning capabilities
- decision theoretic
- agent learns
- multiple agents
- agent model
- agent architecture
- mobile agents
- reinforcement learning algorithms
- reinforcement learning agents
- exploration strategy
- dynamic programming
- human experts
- action space
- computer chess
- markov decision processes
- learning process
- objective function
- robocup soccer
- balancing exploration and exploitation
- learning algorithm
- model free
- computer programs
- optimal policy
- cooperative
- knowledge base