"A Fully Problem-Dependent Regret Lower Bound for Finite-Horizon MDPs."

Andrea Tirinzoni, Matteo Pirotta, Alessandro Lazaric (2021)

Details and statistics

DOI:

access: open

type: Informal or Other Publication

metadata version: 2021-06-30

a service of  Schloss Dagstuhl - Leibniz Center for Informatics