Robot reinforcement learning using EEG-based reward signals.
Iñaki IturrateLuis MontesanoJavier MinguezPublished in: ICRA (2010)
Keyphrases
- reinforcement learning
- eeg signals
- signal processing
- real robot
- perceptual aliasing
- robot control
- mobile robot
- eeg data
- brain activity
- physiological signals
- function approximation
- state space
- control signals
- single channel
- signal acquisition
- sensory inputs
- brain signals
- robotic systems
- autonomous robots
- eligibility traces
- vision system
- sleep stage
- reward function
- human robot interaction
- reinforcement learning algorithms
- temporal difference
- optimal policy
- machine learning
- robot navigation
- brain computer interface
- learning agent
- autonomous learning
- path planning
- action selection
- humanoid robot
- partially observable environments
- independent component analysis
- learning algorithm
- sound source
- multi robot
- reinforcement learning methods
- model free
- frequency ranges
- motor imagery
- agent learns
- learning capabilities
- reward shaping
- markov decision process
- markov decision processes
- reward signal
- optimal control
- formation control
- single trial
- average reward
- policy gradient
- healthy subjects