Brain-Inspired Emergence of Behaviors in Mobile Robots by Reinforcement Learning with Internal Rewards.
Masumi IshikawaTakao HagiwaraNaoyuki YamamotoFumiko KiriakePublished in: HIS (2008)
Keyphrases
- reinforcement learning
- mobile robot
- real robot
- sensory inputs
- external world
- unknown environments
- markov decision processes
- path planning
- state space
- function approximation
- obstacle avoidance
- indoor environments
- model free
- robot control
- human brain
- motion control
- autonomous robots
- dynamic environments
- reinforcement learning algorithms
- mobile robotics
- learning algorithm
- human behavior
- machine learning
- motion planning
- optimal policy
- action selection
- reward shaping
- multi agent
- outdoor environments
- temporal difference
- cognitive states
- function approximators
- brain images
- action space
- markov decision process
- brain activity
- partially observable
- dynamic programming
- learning process
- eeg signals
- reward function