Truncating Temporal Differences: On the Efficient Implementation of TD(lambda) for Reinforcement Learning.
Pawel CichoszPublished in: J. Artif. Intell. Res. (1995)
Keyphrases
- temporal difference
- efficient implementation
- reinforcement learning
- td learning
- function approximation
- reinforcement learning algorithms
- model free
- evaluation function
- temporal difference learning
- monte carlo
- policy evaluation
- step size
- active set
- function approximators
- policy iteration
- state space
- action selection
- temporal difference methods
- supervised learning
- dynamic programming
- markov decision processes
- reinforcement learning methods
- learning algorithm
- optimal policy
- eligibility traces
- linear combination
- reinforcement learning problems
- neural network