Approximate Planning in Large POMDPs via Reusable Trajectories

@inproceedings{Kearns1999ApproximatePI,
  title={Approximate Planning in Large POMDPs via Reusable Trajectories},
  author={Michael Kearns and Yishay Mansour and Andrew Y. Ng},
  booktitle={NIPS},
  year={1999}
}
We considertheproblemof reliably choosinga near-beststrategy from a restrictedclassof strategies in a partially observableMarkov decision process(POMDP).We assumewe are given the ability to simulatethe POMDP, andstudywhatmight becalledthesamplecomplexity — thatis, theamountof dataonemustgeneratein thePOMDPin orderto choosea goodstrategy. We prove upperboundson thesamplecomplexity showing that,evenfor infinitelylargeandarbitrarily complex POMDPs,theamount of dataneededcanbefinite,anddependsonly… CONTINUE READING
Highly Influential
This paper has highly influenced 15 other papers. REVIEW HIGHLY INFLUENTIAL CITATIONS
Highly Cited
This paper has 195 citations. REVIEW CITATIONS

Topic

Statistics

01020'00'02'04'06'08'10'12'14'16'18
Citations per Year

196 Citations

Semantic Scholar estimates that this publication has 196 citations based on the available data.

See our FAQ for additional information.