Home > Publications . Search All . Browse All . Country . Browse PSC Pubs . PSC Report Series

PSC In The News

RSS Feed icon

Levy says ACA has helped increase rates of insured, but rates still lowest among poor

Bruch reveals key decision criteria in making first cuts on dating sites

Murphy on extending health support via a smart phone and JITAI

More News

Highlights

U-M ranked #4 in USN&WR's top public universities

Frey's new report explores how the changing US electorate could shape the next 5 presidential elections, 2016 to 2032

U-M's Data Science Initiative offers expanded consulting services via CSCAR

Elizabeth Bruch promoted to Associate Professor

Next Brown Bag

Mon, Oct 3 at noon:
Longevity, Education, & Income, Hoyt Bleakley

Sampling strategies for batch mode reinforcement learning

Publication Abstract

Fonteneau, Raphael, Susan A. Murphy, L. Wehenkel, and D. Ernst. 2013. "Sampling strategies for batch mode reinforcement learning." Revue d'Intelligence Artificielle, 27(2): 171-194.

We propose two strategies for experiment selection in the context of batch mode reinforcement learning. The first strategy is based on the idea that the most interesting experiments to carry out at some stage are those that are the most liable to falsify the current hypothesis about the optimal control policy. We cast this idea in a context where a policy learning algorithm and a model identification method are given a priori. The second strategy exploits recently published methods for computing bounds on the return of control policies from a set of trajectories in order to sample the state-action space so as to be able to discriminate between optimal and non-optimal policies. Both strategies are experimentally validated, showing promising results. © 2013 Lavoisier.

DOI:10.3166/RIA.27.171-194 (Full Text)

Browse | Search : All Pubs | Next