Robotics and the Q-analysis of behaviour

Pejman Iravani, J H Johnson, L Rapanotti

Research output: Contribution to conferencePaper

Abstract

A new method is presented for the analysis of complex multi-agent robotic behaviours, based on observing behaviour and abstracting models on which to base control. This involves (i) building an appropriate representation of the scene using the `best' features, (ii) classifying each scene by its features, and (iii) learning the correlation between scene classes and the actions performed in each. A novel aspect is the use of Q-analysis to investigate relational structure between many possible observable features, and congurations in the scene. Q-analysis helps in the selection of appropriate features and is the basis of our combinatorial classification. The methods and algorithms presented are experimentally validated using data from the RoboCup simulation league. We conclude that Q-analysis could be a powerful new approach in robot control.

Conference

Conference10th International Symposium on Artificial Life and Robotics (AROB)
CountryJapan
CityOita
Period4/02/056/02/05

Fingerprint

Robotics
Robots

Cite this

Iravani, P., Johnson, J. H., & Rapanotti, L. (2005). Robotics and the Q-analysis of behaviour. Paper presented at 10th International Symposium on Artificial Life and Robotics (AROB), Oita, Japan.

Robotics and the Q-analysis of behaviour. / Iravani, Pejman; Johnson, J H; Rapanotti, L.

2005. Paper presented at 10th International Symposium on Artificial Life and Robotics (AROB), Oita, Japan.

Research output: Contribution to conferencePaper

Iravani, P, Johnson, JH & Rapanotti, L 2005, 'Robotics and the Q-analysis of behaviour' Paper presented at 10th International Symposium on Artificial Life and Robotics (AROB), Oita, Japan, 4/02/05 - 6/02/05, .
Iravani P, Johnson JH, Rapanotti L. Robotics and the Q-analysis of behaviour. 2005. Paper presented at 10th International Symposium on Artificial Life and Robotics (AROB), Oita, Japan.
Iravani, Pejman ; Johnson, J H ; Rapanotti, L. / Robotics and the Q-analysis of behaviour. Paper presented at 10th International Symposium on Artificial Life and Robotics (AROB), Oita, Japan.
@conference{491474174ba547979de6309f6c82b9dc,
title = "Robotics and the Q-analysis of behaviour",
abstract = "A new method is presented for the analysis of complex multi-agent robotic behaviours, based on observing behaviour and abstracting models on which to base control. This involves (i) building an appropriate representation of the scene using the `best' features, (ii) classifying each scene by its features, and (iii) learning the correlation between scene classes and the actions performed in each. A novel aspect is the use of Q-analysis to investigate relational structure between many possible observable features, and congurations in the scene. Q-analysis helps in the selection of appropriate features and is the basis of our combinatorial classification. The methods and algorithms presented are experimentally validated using data from the RoboCup simulation league. We conclude that Q-analysis could be a powerful new approach in robot control.",
author = "Pejman Iravani and Johnson, {J H} and L Rapanotti",
year = "2005",
language = "English",
note = "10th International Symposium on Artificial Life and Robotics (AROB) ; Conference date: 04-02-2005 Through 06-02-2005",

}

TY - CONF

T1 - Robotics and the Q-analysis of behaviour

AU - Iravani, Pejman

AU - Johnson, J H

AU - Rapanotti, L

PY - 2005

Y1 - 2005

N2 - A new method is presented for the analysis of complex multi-agent robotic behaviours, based on observing behaviour and abstracting models on which to base control. This involves (i) building an appropriate representation of the scene using the `best' features, (ii) classifying each scene by its features, and (iii) learning the correlation between scene classes and the actions performed in each. A novel aspect is the use of Q-analysis to investigate relational structure between many possible observable features, and congurations in the scene. Q-analysis helps in the selection of appropriate features and is the basis of our combinatorial classification. The methods and algorithms presented are experimentally validated using data from the RoboCup simulation league. We conclude that Q-analysis could be a powerful new approach in robot control.

AB - A new method is presented for the analysis of complex multi-agent robotic behaviours, based on observing behaviour and abstracting models on which to base control. This involves (i) building an appropriate representation of the scene using the `best' features, (ii) classifying each scene by its features, and (iii) learning the correlation between scene classes and the actions performed in each. A novel aspect is the use of Q-analysis to investigate relational structure between many possible observable features, and congurations in the scene. Q-analysis helps in the selection of appropriate features and is the basis of our combinatorial classification. The methods and algorithms presented are experimentally validated using data from the RoboCup simulation league. We conclude that Q-analysis could be a powerful new approach in robot control.

M3 - Paper

ER -