Compressed kernelized least squares temporal difference learning via random projections

Aiwen Jiang, Bo Liu, Shuai Li

Research output: Journal article publicationJournal articleAcademic researchpeer-review


We propose a compressed kernelized least squares temporal difference learning (CKLSTD) algorithm for reinforcement learning in large state space by incorporate kernel trick and random projection into traditional LSTD. Our proposed method can be viewed as kernelized version of LSTD-RP and compressed version of KLSTD, wherein the proposed method preserves the merits of both of them. The experimental results suggest that our proposed method can gain better performance in high dimensional feature space.
Original languageEnglish
Pages (from-to)8955-8962
Number of pages8
JournalJournal of Computational Information Systems
Issue number22
Publication statusPublished - 15 Nov 2013
Externally publishedYes


  • Compressive regression
  • Kernel sparsification
  • Least squares temporal difference learning
  • Random projects

ASJC Scopus subject areas

  • Information Systems
  • Computer Science Applications

Cite this