Using classification learning in companion modeling

Daisuke Torii, Francois Bousquet, Toru Ishida, Guy Trébuil, Chirawat Vejpas

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Companion Modeling is a methodology used to facilitate adaptive management of renewable resources by their users. It is using role-playing games (RPG) and multiagent simulations to validate initial models representing the functioning of complex systems to be managed. In this research, we propose a novel agent model construction methodology in which classification learning is applied to the RPG log data in Companion Modeling. This methodology enables a systematic model construction that handles multi-parameters, independent of the modelers' ability. There are three problems in applying classification learning to the RPG log data: 1) It is difficult to gather enough data for the number of features because the cost of gathering data is high. 2) Noise data can affect the learning results because the amount of data may be insufficient. 3) The learning results should be explained as a human decision making model and should be recognized by the expert as reflecting reality. We realized an agent model construction system using the following two approaches: 1) Using a feature selection method, the feature subset that has the best prediction accuracy is identified. In this process, the important features chosen by the expert are always included. 2) The expert eliminates irrelevant features from the learning results after evaluating the learning model through a visualization of the results. Finally, using the RPG log data from a Companion Modeling case study on rice production in northeastern Thailand, we confirm the capability of this methodology.

Original languageEnglish
Title of host publicationMulti-Agent Systems for Society - 8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005, Revised Selected Papers
Pages255-269
Number of pages15
DOIs
Publication statusPublished - 2009 Nov 9
Externally publishedYes
Event8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005 - Kuala Lumpur, Malaysia
Duration: 2005 Sep 262005 Sep 28

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4078 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005
CountryMalaysia
CityKuala Lumpur
Period05/9/2605/9/28

Fingerprint

Datalog
Modeling
Game
Methodology
Model
Renewable Resources
Multi-agent Simulation
Set theory
Feature Selection
Learning
Large scale systems
Feature extraction
Complex Systems
Eliminate
Visualization
Decision making
Decision Making
Subset
Prediction
Costs

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Torii, D., Bousquet, F., Ishida, T., Trébuil, G., & Vejpas, C. (2009). Using classification learning in companion modeling. In Multi-Agent Systems for Society - 8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005, Revised Selected Papers (pp. 255-269). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4078 LNAI). https://doi.org/10.1007/978-3-642-03339-1_21

Using classification learning in companion modeling. / Torii, Daisuke; Bousquet, Francois; Ishida, Toru; Trébuil, Guy; Vejpas, Chirawat.

Multi-Agent Systems for Society - 8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005, Revised Selected Papers. 2009. p. 255-269 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4078 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Torii, D, Bousquet, F, Ishida, T, Trébuil, G & Vejpas, C 2009, Using classification learning in companion modeling. in Multi-Agent Systems for Society - 8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4078 LNAI, pp. 255-269, 8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005, Kuala Lumpur, Malaysia, 05/9/26. https://doi.org/10.1007/978-3-642-03339-1_21
Torii D, Bousquet F, Ishida T, Trébuil G, Vejpas C. Using classification learning in companion modeling. In Multi-Agent Systems for Society - 8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005, Revised Selected Papers. 2009. p. 255-269. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-03339-1_21
Torii, Daisuke ; Bousquet, Francois ; Ishida, Toru ; Trébuil, Guy ; Vejpas, Chirawat. / Using classification learning in companion modeling. Multi-Agent Systems for Society - 8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005, Revised Selected Papers. 2009. pp. 255-269 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{71148da827e64bfcaea93390889e4fd1,
title = "Using classification learning in companion modeling",
abstract = "Companion Modeling is a methodology used to facilitate adaptive management of renewable resources by their users. It is using role-playing games (RPG) and multiagent simulations to validate initial models representing the functioning of complex systems to be managed. In this research, we propose a novel agent model construction methodology in which classification learning is applied to the RPG log data in Companion Modeling. This methodology enables a systematic model construction that handles multi-parameters, independent of the modelers' ability. There are three problems in applying classification learning to the RPG log data: 1) It is difficult to gather enough data for the number of features because the cost of gathering data is high. 2) Noise data can affect the learning results because the amount of data may be insufficient. 3) The learning results should be explained as a human decision making model and should be recognized by the expert as reflecting reality. We realized an agent model construction system using the following two approaches: 1) Using a feature selection method, the feature subset that has the best prediction accuracy is identified. In this process, the important features chosen by the expert are always included. 2) The expert eliminates irrelevant features from the learning results after evaluating the learning model through a visualization of the results. Finally, using the RPG log data from a Companion Modeling case study on rice production in northeastern Thailand, we confirm the capability of this methodology.",
author = "Daisuke Torii and Francois Bousquet and Toru Ishida and Guy Tr{\'e}buil and Chirawat Vejpas",
year = "2009",
month = "11",
day = "9",
doi = "10.1007/978-3-642-03339-1_21",
language = "English",
isbn = "3642033377",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "255--269",
booktitle = "Multi-Agent Systems for Society - 8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005, Revised Selected Papers",

}

TY - GEN

T1 - Using classification learning in companion modeling

AU - Torii, Daisuke

AU - Bousquet, Francois

AU - Ishida, Toru

AU - Trébuil, Guy

AU - Vejpas, Chirawat

PY - 2009/11/9

Y1 - 2009/11/9

N2 - Companion Modeling is a methodology used to facilitate adaptive management of renewable resources by their users. It is using role-playing games (RPG) and multiagent simulations to validate initial models representing the functioning of complex systems to be managed. In this research, we propose a novel agent model construction methodology in which classification learning is applied to the RPG log data in Companion Modeling. This methodology enables a systematic model construction that handles multi-parameters, independent of the modelers' ability. There are three problems in applying classification learning to the RPG log data: 1) It is difficult to gather enough data for the number of features because the cost of gathering data is high. 2) Noise data can affect the learning results because the amount of data may be insufficient. 3) The learning results should be explained as a human decision making model and should be recognized by the expert as reflecting reality. We realized an agent model construction system using the following two approaches: 1) Using a feature selection method, the feature subset that has the best prediction accuracy is identified. In this process, the important features chosen by the expert are always included. 2) The expert eliminates irrelevant features from the learning results after evaluating the learning model through a visualization of the results. Finally, using the RPG log data from a Companion Modeling case study on rice production in northeastern Thailand, we confirm the capability of this methodology.

AB - Companion Modeling is a methodology used to facilitate adaptive management of renewable resources by their users. It is using role-playing games (RPG) and multiagent simulations to validate initial models representing the functioning of complex systems to be managed. In this research, we propose a novel agent model construction methodology in which classification learning is applied to the RPG log data in Companion Modeling. This methodology enables a systematic model construction that handles multi-parameters, independent of the modelers' ability. There are three problems in applying classification learning to the RPG log data: 1) It is difficult to gather enough data for the number of features because the cost of gathering data is high. 2) Noise data can affect the learning results because the amount of data may be insufficient. 3) The learning results should be explained as a human decision making model and should be recognized by the expert as reflecting reality. We realized an agent model construction system using the following two approaches: 1) Using a feature selection method, the feature subset that has the best prediction accuracy is identified. In this process, the important features chosen by the expert are always included. 2) The expert eliminates irrelevant features from the learning results after evaluating the learning model through a visualization of the results. Finally, using the RPG log data from a Companion Modeling case study on rice production in northeastern Thailand, we confirm the capability of this methodology.

UR - http://www.scopus.com/inward/record.url?scp=70350646549&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70350646549&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-03339-1_21

DO - 10.1007/978-3-642-03339-1_21

M3 - Conference contribution

AN - SCOPUS:70350646549

SN - 3642033377

SN - 9783642033377

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 255

EP - 269

BT - Multi-Agent Systems for Society - 8th Pacific Rim International Workshop on Multi-Agents, PRIMA 2005, Revised Selected Papers

ER -