Modular Neural Networks for Reinforcement Learning with Temporal Intrinsic Rewards.
Johane TakeuchiOsamu ShounoHiroshi TsujinoPublished in: IJCNN (2007)
Keyphrases
- reinforcement learning
- modular neural networks
- evolutionary learning
- markov decision processes
- bio inspired
- function approximation
- state space
- learning algorithm
- hybrid intelligent systems
- reinforcement learning algorithms
- model free
- temporal information
- reward function
- pattern recognition
- reward shaping
- optimal policy
- machine learning
- dynamic programming
- multi agent
- genetic algorithm
- control system