Imperial College London

DrPetarKormushev

Faculty of EngineeringDyson School of Design Engineering

Lecturer
 
 
 
//

Contact

 

+44 (0)20 7594 9235p.kormushev Website

 
 
//

Location

 

10-12 Prince's GardensSouth Kensington Campus

//

Summary

 

Publications

Citation

BibTex format

@inproceedings{Kormushev:2012,
author = {Kormushev, P and Caldwell, DG},
title = {Direct policy search reinforcement learning based on particle filtering},
url = {http://kormushev.com/papers/Kormushev_EWRL-2012.pdf},
year = {2012}
}

RIS format (EndNote, RefMan)

TY  - CPAPER
AB - We reveal a link between particle filtering methods and direct policy search reinforcementlearning, and propose a novel reinforcement learning algorithm, based heavily on ideasborrowed from particle filters. A major advantage of the proposed algorithm is its ability toperform global search in policy space and thus find the globally optimal policy. We validatethe approach on one- and two-dimensional problems with multiple optima, and compareits performance to a global random sampling method, and a state-of-the-art ExpectationMaximizationbased reinforcement learning algorithm.
AU - Kormushev,P
AU - Caldwell,DG
PY - 2012///
TI - Direct policy search reinforcement learning based on particle filtering
UR - http://kormushev.com/papers/Kormushev_EWRL-2012.pdf
UR - http://hdl.handle.net/10044/1/26091
ER -