Reward Function and Initial Values: Better Choices for Accelerated Goal-Directed Reinforcement Learning.
Laëtitia MatignonGuillaume J. LaurentNadine Le Fort-PiatPublished in: ICANN (1) (2006)
Keyphrases
- goal directed
- reward function
- initial values
- reinforcement learning
- reinforcement learning algorithms
- markov decision processes
- state space
- partially observable
- optimal policy
- inverse reinforcement learning
- fuzzy model
- information granulation
- markov decision process
- multiple agents
- autonomous robots
- membership functions
- autonomous learning
- multi agent
- transition model
- model free
- machine learning
- initially unknown
- temporal difference
- transition probabilities
- learning algorithm
- state variables
- function approximation
- average reward
- artificial intelligence
- neural network
- action selection
- control policies
- mobile robot
- dynamic programming
- clustering algorithm
- hierarchical reinforcement learning