Attractor Neural States: A Brain-Inspired Complementary Approach to Reinforcement Learning.
Oussama H. HamidJochen BraunPublished in: IJCCI (2017)
Keyphrases
- reinforcement learning
- sensory inputs
- perceptual aliasing
- computational neuroscience
- fitted q iteration
- biologically plausible
- functional connectivity
- cellular automata
- spike trains
- state abstraction
- network architecture
- neural network
- transition model
- function approximation
- basal ganglia
- state space
- state action
- markov decision problems
- event related potentials
- action selection
- fixed point
- human brain
- biologically inspired
- model free
- neural model
- learning algorithm
- state variables
- dynamic behavior
- primary visual cortex
- motor control
- neural mechanisms
- phase space
- function approximators
- sensory motor
- associative memory
- markov decision processes
- brain images
- working memory
- electrical activity
- sensory data
- functional magnetic resonance imaging
- initial state
- robot control
- partially observable
- brain computer interface
- optimal control
- feed forward
- dynamical systems
- computational intelligence
- multi agent
- machine learning