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

@unpublished{Kormushev:2009,
author = {Kormushev, P and Nomoto, K and Dong, F and Hirota, K},
title = {Eligibility Propagation to Speed up Time Hopping for Reinforcement Learning},
url = {http://arxiv.org/abs/0904.0546v1},
year = {2009}
}

RIS format (EndNote, RefMan)

TY  - UNPB
AB - A mechanism called Eligibility Propagation is proposed to speed up the TimeHopping technique used for faster Reinforcement Learning in simulations.Eligibility Propagation provides for Time Hopping similar abilities to whateligibility traces provide for conventional Reinforcement Learning. Itpropagates values from one state to all of its temporal predecessors using astate transitions graph. Experiments on a simulated biped crawling robotconfirm that Eligibility Propagation accelerates the learning process more than3 times.
AU - Kormushev,P
AU - Nomoto,K
AU - Dong,F
AU - Hirota,K
PY - 2009///
TI - Eligibility Propagation to Speed up Time Hopping for Reinforcement Learning
UR - http://arxiv.org/abs/0904.0546v1
UR - http://hdl.handle.net/10044/1/26076
ER -