Enhancing metacognitive reinforcement learning using reward structures and feedback.
Paul M. KruegerFalk LiederTom GriffithsPublished in: CogSci (2017)
Keyphrases
- reinforcement learning
- state space
- eligibility traces
- model free
- markov decision processes
- reward function
- reinforcement learning algorithms
- temporal difference
- function approximation
- optimal policy
- policy gradient
- dynamic programming
- learning environment
- learning problems
- learning algorithm
- optimal control
- partially observable environments
- average reward
- reward signal
- supervised learning
- relevance feedback
- machine learning
- action selection
- learning process
- multi agent
- multi agent reinforcement learning
- feedback mechanisms
- information retrieval
- neural network