Improving reinforcement learning results with qualitative spatial representation

Thiago Pedro Donadon Homem, Danilo Hernani Perico, Paulo Eduardo Santos, Anna Helena Reali Costa, Reinaldo Augusto Da Costa Bianchi

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Citations (Scopus)

Abstract

Reinforcement learning and Qualitative Spatial Reasoning methods have been successfully applied to create agents able to solve Artificial Intelligence problems in games, robotics, simulated or real. Generally, reinforcement learning methods represent the objects' position as quantitative values, performing the experiments considering these values. However, the humancommonsense understanding of the world is qualitative. This work proposes a hybrid method, that uses a qualitative formalism with reinforcement learning, named QRL, and is able to get better results than traditional methods. We have applied this proposal in the robot soccer domain and compared the results with traditional reinforcement learning method. The results show that, by using a qualitative spatial representation with reinforcement learning, the agent can learn optimal policies and perform more goals than quantitative representation.

Original languageEnglish
Title of host publicationProceedings - 2017 Brazilian Conference on Intelligent Systems, BRACIS 2017
PublisherInstitute of Electrical and Electronics Engineers
Pages151-156
Number of pages6
ISBN (Electronic)9781538624074
DOIs
Publication statusPublished - 28 Jun 2017
Externally publishedYes
Event6th Brazilian Conference on Intelligent Systems, BRACIS 2017 - Uberlandia, Brazil
Duration: 2 Oct 20175 Oct 2017

Publication series

NameProceedings - 2017 Brazilian Conference on Intelligent Systems, BRACIS 2017
Volume2018-January

Conference

Conference6th Brazilian Conference on Intelligent Systems, BRACIS 2017
Country/TerritoryBrazil
CityUberlandia
Period2/10/175/10/17

Keywords

  • qualitative spatial reasoning
  • reinforcement learning
  • robot-soccer

Fingerprint

Dive into the research topics of 'Improving reinforcement learning results with qualitative spatial representation'. Together they form a unique fingerprint.

Cite this