Evolving an autonomous agent for non-Markovian reinforcement learning.
Jae-Yoon JungJames A. ReggiaPublished in: GECCO (2009)
Keyphrases
- autonomous agents
- reinforcement learning
- multi agent
- reinforcement learning agents
- dynamic environments
- function approximation
- multi agent environments
- multi agent systems
- reward signal
- multiagent systems
- control architecture
- learning algorithm
- temporal difference
- complex environments
- model free
- state space
- changing environment
- action selection
- reward function
- reinforcement learning algorithms
- markov decision process
- machine learning
- markov decision processes
- partially observable domains
- hidden state
- optimal policy
- transfer learning
- learning capabilities
- bdi agents
- reinforcement learning methods
- computational systems
- dynamic programming
- decision processes
- search space
- optimal control