The Optimal Approximation Factors in Misspecified Off-Policy Value Function Estimation.
Philip AmortilaNan JiangCsaba SzepesváriPublished in: CoRR (2023)
Keyphrases
- piecewise linear
- lp norm
- closed form
- dynamic programming
- estimation error
- factors that influence
- accurate estimation
- continuous functions
- control policy
- approximation algorithms
- closed form solutions
- optimal solution
- convex functions
- key factors
- optimal design
- error tolerance
- approximation error
- factors that affect
- series expansion
- data sets
- pointwise
- estimation algorithm
- optimal control
- worst case
- reinforcement learning
- genetic algorithm