Provably Efficient Offline Reinforcement Learning with Perturbed Data Sources.

Chengshuai Shi Wei Xiong Cong Shen Jing Yang

Published in: ICML (2023)

Keyphrases

data sources
reinforcement learning
real world
learning algorithm
databases
data model
markov decision processes
genetic algorithm
function approximation
computationally expensive
cost effective
multiple sources
data integration
computationally efficient
worst case
state space
case study
information systems