Extension of Companion Modeling Using Classification Learning

Daisuke Torii, François Bousquet, Toru Ishida

Research output: Contribution to journalArticle

Abstract

Companion Modeling is a methodology of refining initial models for understanding reality through a role-playing game (RPG) and a multiagent simulation. 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 being the result that reflects reality. We realized an agent model construction system using the following two approaches: 1. Using a feature selction 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 the Companion Modeling of agricultural economics in northeastern Thailand, we confirm the capability of this methodology. keywords: multiagent-based simulation, modeling, machine learning, classification learning, feature selection.

Original languageEnglish
Pages (from-to)379-386
Number of pages8
JournalTransactions of the Japanese Society for Artificial Intelligence
Volume20
DOIs
Publication statusPublished - 2005 Jan 1
Externally publishedYes

Fingerprint

Set theory
Refining
Learning systems
Feature extraction
Visualization
Decision making
Economics
Computer simulation
Costs

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this

Extension of Companion Modeling Using Classification Learning. / Torii, Daisuke; Bousquet, François; Ishida, Toru.

In: Transactions of the Japanese Society for Artificial Intelligence, Vol. 20, 01.01.2005, p. 379-386.

Research output: Contribution to journalArticle

@article{9daf90c109824e6aa1eff54ad4877f20,
title = "Extension of Companion Modeling Using Classification Learning",
abstract = "Companion Modeling is a methodology of refining initial models for understanding reality through a role-playing game (RPG) and a multiagent simulation. 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 being the result that reflects reality. We realized an agent model construction system using the following two approaches: 1. Using a feature selction 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 the Companion Modeling of agricultural economics in northeastern Thailand, we confirm the capability of this methodology. keywords: multiagent-based simulation, modeling, machine learning, classification learning, feature selection.",
author = "Daisuke Torii and Fran{\cc}ois Bousquet and Toru Ishida",
year = "2005",
month = "1",
day = "1",
doi = "10.1527/tjsai.20.379",
language = "English",
volume = "20",
pages = "379--386",
journal = "Transactions of the Japanese Society for Artificial Intelligence",
issn = "1346-0714",
publisher = "Japanese Society for Artificial Intelligence",

}

TY - JOUR

T1 - Extension of Companion Modeling Using Classification Learning

AU - Torii, Daisuke

AU - Bousquet, François

AU - Ishida, Toru

PY - 2005/1/1

Y1 - 2005/1/1

N2 - Companion Modeling is a methodology of refining initial models for understanding reality through a role-playing game (RPG) and a multiagent simulation. 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 being the result that reflects reality. We realized an agent model construction system using the following two approaches: 1. Using a feature selction 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 the Companion Modeling of agricultural economics in northeastern Thailand, we confirm the capability of this methodology. keywords: multiagent-based simulation, modeling, machine learning, classification learning, feature selection.

AB - Companion Modeling is a methodology of refining initial models for understanding reality through a role-playing game (RPG) and a multiagent simulation. 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 being the result that reflects reality. We realized an agent model construction system using the following two approaches: 1. Using a feature selction 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 the Companion Modeling of agricultural economics in northeastern Thailand, we confirm the capability of this methodology. keywords: multiagent-based simulation, modeling, machine learning, classification learning, feature selection.

UR - http://www.scopus.com/inward/record.url?scp=85024740031&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85024740031&partnerID=8YFLogxK

U2 - 10.1527/tjsai.20.379

DO - 10.1527/tjsai.20.379

M3 - Article

AN - SCOPUS:85024740031

VL - 20

SP - 379

EP - 386

JO - Transactions of the Japanese Society for Artificial Intelligence

JF - Transactions of the Japanese Society for Artificial Intelligence

SN - 1346-0714

ER -