Estimation and Approximation Bounds for Gradient-Based Reinforcement Learning.
Peter L. BartlettJonathan BaxterPublished in: COLT (2000)
Keyphrases
- reinforcement learning
- error bounds
- upper bound
- model free
- lower bound
- reinforcement learning algorithms
- markovian decision
- estimation accuracy
- function approximation
- stage stochastic programs
- state space
- error tolerance
- approximation methods
- optimal policy
- markov decision processes
- closed form
- average case
- multi agent
- approximation algorithms
- worst case
- learning process
- estimation algorithm
- upper and lower bounds
- density estimation
- accurate estimation
- parameter estimation
- importance sampling
- approximation error
- maximum likelihood estimator
- supervised learning
- dynamic programming