Improving virtual sample generation for small sample learning with dependent attributes

Liang Sian Lin, Der Chiang Li, Chih Wei Pan

研究成果: Conference contribution

1 引文 (Scopus)

摘要

Since the product life cycles are getting shorter and shorter, the issue of small data set learning has drawn more and more attentions in both academics and enterprises. Many methods have been proposed to improve the learning performance of small data set. In these methods, the virtual sample generation approach is the most popular technique for improving small data learning. In the process of virtual sample generation, the attribute independence in small data is the key part to determine the learning performance, because it is the necessary assumption before generating virtual samples. However, in the real world, attributes in the data set usually are not mutual independent. Therefore, this paper proposes a new process to generate independent virtual samples based on the box-and-whisker plot domain estimation. In order to validate the effectiveness of the proposed method, one data set is used to calculate the classification accuracy average and standard deviation based on the support vector machine. The results of the experiment show that the presented method has a superior classification performance than other methods.

原文English
主出版物標題Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016
編輯Ayako Hiramatsu, Tokuro Matsuo, Akimitsu Kanzaki, Norihisa Komoda
發行者Institute of Electrical and Electronics Engineers Inc.
頁面715-718
頁數4
ISBN(電子)9781467389853
DOIs
出版狀態Published - 2016 八月 31
事件5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016 - Kumamoto, Japan
持續時間: 2016 七月 102016 七月 14

出版系列

名字Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016

Other

Other5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016
國家Japan
城市Kumamoto
期間16-07-1016-07-14

指紋

Support vector machines
Life cycle
Industry
Experiments

All Science Journal Classification (ASJC) codes

  • Information Systems
  • Computer Networks and Communications
  • Computer Science Applications
  • Computer Vision and Pattern Recognition

引用此文

Lin, L. S., Li, D. C., & Pan, C. W. (2016). Improving virtual sample generation for small sample learning with dependent attributes. 於 A. Hiramatsu, T. Matsuo, A. Kanzaki, & N. Komoda (編輯), Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016 (頁 715-718). [7557704] (Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IIAI-AAI.2016.18
Lin, Liang Sian ; Li, Der Chiang ; Pan, Chih Wei. / Improving virtual sample generation for small sample learning with dependent attributes. Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016. 編輯 / Ayako Hiramatsu ; Tokuro Matsuo ; Akimitsu Kanzaki ; Norihisa Komoda. Institute of Electrical and Electronics Engineers Inc., 2016. 頁 715-718 (Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016).
@inproceedings{3b869566bd084d42ba1480793f666f7d,
title = "Improving virtual sample generation for small sample learning with dependent attributes",
abstract = "Since the product life cycles are getting shorter and shorter, the issue of small data set learning has drawn more and more attentions in both academics and enterprises. Many methods have been proposed to improve the learning performance of small data set. In these methods, the virtual sample generation approach is the most popular technique for improving small data learning. In the process of virtual sample generation, the attribute independence in small data is the key part to determine the learning performance, because it is the necessary assumption before generating virtual samples. However, in the real world, attributes in the data set usually are not mutual independent. Therefore, this paper proposes a new process to generate independent virtual samples based on the box-and-whisker plot domain estimation. In order to validate the effectiveness of the proposed method, one data set is used to calculate the classification accuracy average and standard deviation based on the support vector machine. The results of the experiment show that the presented method has a superior classification performance than other methods.",
author = "Lin, {Liang Sian} and Li, {Der Chiang} and Pan, {Chih Wei}",
year = "2016",
month = "8",
day = "31",
doi = "10.1109/IIAI-AAI.2016.18",
language = "English",
series = "Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "715--718",
editor = "Ayako Hiramatsu and Tokuro Matsuo and Akimitsu Kanzaki and Norihisa Komoda",
booktitle = "Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016",
address = "United States",

}

Lin, LS, Li, DC & Pan, CW 2016, Improving virtual sample generation for small sample learning with dependent attributes. 於 A Hiramatsu, T Matsuo, A Kanzaki & N Komoda (編輯), Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016., 7557704, Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016, Institute of Electrical and Electronics Engineers Inc., 頁 715-718, 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016, Kumamoto, Japan, 16-07-10. https://doi.org/10.1109/IIAI-AAI.2016.18

Improving virtual sample generation for small sample learning with dependent attributes. / Lin, Liang Sian; Li, Der Chiang; Pan, Chih Wei.

Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016. 編輯 / Ayako Hiramatsu; Tokuro Matsuo; Akimitsu Kanzaki; Norihisa Komoda. Institute of Electrical and Electronics Engineers Inc., 2016. p. 715-718 7557704 (Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016).

研究成果: Conference contribution

TY - GEN

T1 - Improving virtual sample generation for small sample learning with dependent attributes

AU - Lin, Liang Sian

AU - Li, Der Chiang

AU - Pan, Chih Wei

PY - 2016/8/31

Y1 - 2016/8/31

N2 - Since the product life cycles are getting shorter and shorter, the issue of small data set learning has drawn more and more attentions in both academics and enterprises. Many methods have been proposed to improve the learning performance of small data set. In these methods, the virtual sample generation approach is the most popular technique for improving small data learning. In the process of virtual sample generation, the attribute independence in small data is the key part to determine the learning performance, because it is the necessary assumption before generating virtual samples. However, in the real world, attributes in the data set usually are not mutual independent. Therefore, this paper proposes a new process to generate independent virtual samples based on the box-and-whisker plot domain estimation. In order to validate the effectiveness of the proposed method, one data set is used to calculate the classification accuracy average and standard deviation based on the support vector machine. The results of the experiment show that the presented method has a superior classification performance than other methods.

AB - Since the product life cycles are getting shorter and shorter, the issue of small data set learning has drawn more and more attentions in both academics and enterprises. Many methods have been proposed to improve the learning performance of small data set. In these methods, the virtual sample generation approach is the most popular technique for improving small data learning. In the process of virtual sample generation, the attribute independence in small data is the key part to determine the learning performance, because it is the necessary assumption before generating virtual samples. However, in the real world, attributes in the data set usually are not mutual independent. Therefore, this paper proposes a new process to generate independent virtual samples based on the box-and-whisker plot domain estimation. In order to validate the effectiveness of the proposed method, one data set is used to calculate the classification accuracy average and standard deviation based on the support vector machine. The results of the experiment show that the presented method has a superior classification performance than other methods.

UR - http://www.scopus.com/inward/record.url?scp=84988905045&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84988905045&partnerID=8YFLogxK

U2 - 10.1109/IIAI-AAI.2016.18

DO - 10.1109/IIAI-AAI.2016.18

M3 - Conference contribution

AN - SCOPUS:84988905045

T3 - Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016

SP - 715

EP - 718

BT - Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016

A2 - Hiramatsu, Ayako

A2 - Matsuo, Tokuro

A2 - Kanzaki, Akimitsu

A2 - Komoda, Norihisa

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Lin LS, Li DC, Pan CW. Improving virtual sample generation for small sample learning with dependent attributes. 於 Hiramatsu A, Matsuo T, Kanzaki A, Komoda N, 編輯, Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016. Institute of Electrical and Electronics Engineers Inc. 2016. p. 715-718. 7557704. (Proceedings - 2016 5th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2016). https://doi.org/10.1109/IIAI-AAI.2016.18