Paper Search Console

Home Search Page About Contact

Journal Title

Title of Journal: Artif Intell Rev

Search In Journal Title:

Abbravation: Artificial Intelligence Review

Search In Journal Abbravation:

Publisher

Springer Netherlands

Search In Publisher:

DOI

10.1002/chin.198852243

Search In DOI:

ISSN

1573-7462

Search In ISSN:
Search In Title Of Papers:

Research on task decomposition and state abstracti

Authors: Yu Lasheng Jiang Zhongbin Liu Kang
Publish Date: 2011/05/28
Volume: 38, Issue: 2, Pages: 119-127
PDF Link

Abstract

Task decomposition and State abstraction are crucial parts in reinforcement learning It allows an agent to ignore aspects of its current states that are irrelevant to its current decision and therefore speeds up dynamic programming and learning This paper presents the SVI algorithm that uses a dynamic Bayesian network model to construct an influence graph that indicates relationships between state variables SVI performs state abstraction for each subtask by ignoring irrelevant state variables and lower level subtasks Experiment results show that the decomposition of tasks introduced by SVI can significantly accelerate constructing a nearoptimal policy This general framework can be applied to a broad spectrum of complex real world problems such as robotics industrial manufacturing games and others


Keywords:

References


.
Search In Abstract Of Papers:
Other Papers In This Journal:


Search Result: