Contingency training

研究成果: 著書/レポートタイプへの貢献会議での発言

抄録

When applied to high-dimensional datasets, feature selection algorithms might still leave dozens of irrelevant variables in the dataset. Therefore, even after feature selection has been applied, classifiers must be prepared to the presence of irrelevant variables. This paper investigates a new training method called Contingency Training which increases the accuracy as well as the robustness against irrelevant attributes. Contingency training is classifier independent. By subsampling and removing information from each sample, it creates a set of constraints. These constraints aid the method to automatically find proper importance weights of the dataset's features. Experiments are conducted with the contingency training applied to neural networks over traditional datasets as well as datasets with additional irrelevant variables. For all of the tests, contingency training surpassed the unmodified training on datasets with irrelevant variables and even outperformed slightly when only a few or no irrelevant variables were present.

元の言語英語
ホスト出版物のタイトルSICE 2013: International Conference on Instrumentation, Control, Information Technology and System Integration - SICE Annual Conference 2013, Conference Proceedings
ページ1361-1366
ページ数6
出版物ステータス出版済み - 2013
イベント2013 52nd Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2013 - Nagoya, 日本
継続期間: 9 14 20139 17 2013

その他

その他2013 52nd Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2013
日本
Nagoya
期間9/14/139/17/13

Fingerprint

Feature extraction
Classifiers
Neural networks
Experiments

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering
  • Control and Systems Engineering
  • Computer Science Applications

これを引用

Vargas, D. V., Takano, H., & Murata, J. (2013). Contingency training. : SICE 2013: International Conference on Instrumentation, Control, Information Technology and System Integration - SICE Annual Conference 2013, Conference Proceedings (pp. 1361-1366)

Contingency training. / Vargas, Danilo Vasconcellos; Takano, Hirotaka; Murata, Junichi.

SICE 2013: International Conference on Instrumentation, Control, Information Technology and System Integration - SICE Annual Conference 2013, Conference Proceedings. 2013. p. 1361-1366.

研究成果: 著書/レポートタイプへの貢献会議での発言

Vargas, DV, Takano, H & Murata, J 2013, Contingency training. : SICE 2013: International Conference on Instrumentation, Control, Information Technology and System Integration - SICE Annual Conference 2013, Conference Proceedings. pp. 1361-1366, 2013 52nd Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2013, Nagoya, 日本, 9/14/13.
Vargas DV, Takano H, Murata J. Contingency training. : SICE 2013: International Conference on Instrumentation, Control, Information Technology and System Integration - SICE Annual Conference 2013, Conference Proceedings. 2013. p. 1361-1366
Vargas, Danilo Vasconcellos ; Takano, Hirotaka ; Murata, Junichi. / Contingency training. SICE 2013: International Conference on Instrumentation, Control, Information Technology and System Integration - SICE Annual Conference 2013, Conference Proceedings. 2013. pp. 1361-1366
@inproceedings{f0f016be39f5406db1fa15b39b8bc9b7,
title = "Contingency training",
abstract = "When applied to high-dimensional datasets, feature selection algorithms might still leave dozens of irrelevant variables in the dataset. Therefore, even after feature selection has been applied, classifiers must be prepared to the presence of irrelevant variables. This paper investigates a new training method called Contingency Training which increases the accuracy as well as the robustness against irrelevant attributes. Contingency training is classifier independent. By subsampling and removing information from each sample, it creates a set of constraints. These constraints aid the method to automatically find proper importance weights of the dataset's features. Experiments are conducted with the contingency training applied to neural networks over traditional datasets as well as datasets with additional irrelevant variables. For all of the tests, contingency training surpassed the unmodified training on datasets with irrelevant variables and even outperformed slightly when only a few or no irrelevant variables were present.",
author = "Vargas, {Danilo Vasconcellos} and Hirotaka Takano and Junichi Murata",
year = "2013",
language = "English",
pages = "1361--1366",
booktitle = "SICE 2013: International Conference on Instrumentation, Control, Information Technology and System Integration - SICE Annual Conference 2013, Conference Proceedings",

}

TY - GEN

T1 - Contingency training

AU - Vargas, Danilo Vasconcellos

AU - Takano, Hirotaka

AU - Murata, Junichi

PY - 2013

Y1 - 2013

N2 - When applied to high-dimensional datasets, feature selection algorithms might still leave dozens of irrelevant variables in the dataset. Therefore, even after feature selection has been applied, classifiers must be prepared to the presence of irrelevant variables. This paper investigates a new training method called Contingency Training which increases the accuracy as well as the robustness against irrelevant attributes. Contingency training is classifier independent. By subsampling and removing information from each sample, it creates a set of constraints. These constraints aid the method to automatically find proper importance weights of the dataset's features. Experiments are conducted with the contingency training applied to neural networks over traditional datasets as well as datasets with additional irrelevant variables. For all of the tests, contingency training surpassed the unmodified training on datasets with irrelevant variables and even outperformed slightly when only a few or no irrelevant variables were present.

AB - When applied to high-dimensional datasets, feature selection algorithms might still leave dozens of irrelevant variables in the dataset. Therefore, even after feature selection has been applied, classifiers must be prepared to the presence of irrelevant variables. This paper investigates a new training method called Contingency Training which increases the accuracy as well as the robustness against irrelevant attributes. Contingency training is classifier independent. By subsampling and removing information from each sample, it creates a set of constraints. These constraints aid the method to automatically find proper importance weights of the dataset's features. Experiments are conducted with the contingency training applied to neural networks over traditional datasets as well as datasets with additional irrelevant variables. For all of the tests, contingency training surpassed the unmodified training on datasets with irrelevant variables and even outperformed slightly when only a few or no irrelevant variables were present.

UR - http://www.scopus.com/inward/record.url?scp=84888614540&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84888614540&partnerID=8YFLogxK

M3 - Conference contribution

SP - 1361

EP - 1366

BT - SICE 2013: International Conference on Instrumentation, Control, Information Technology and System Integration - SICE Annual Conference 2013, Conference Proceedings

ER -