Dempster-Shafer Fusion of Semi-supervised Learning Methods for Predicting Defaults in Social Lending

Aleum Kim, Sung-Bae Cho

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

3 Citations (Scopus)

Abstract

In social lending, it is hard to know whether borrowers will repay well or not. Most researchers use supervised learning for default prediction, but labeling data by hand is time-consuming. Moreover, labeling results of semi-supervised learning methods are not the same each other. In this paper, we propose a fusion method of label propagation and transductive SVM based on Dempster-Shafer theory for precisely labeling unlabeled data to improve the performance. We remove few unlabeled data with lower reliabilities in labeling results and fusion of the two results based on Dempster-Shafer theory. We have conducted experiments with supervised learning method trained with labeled unlabeled data. As a result, the proposed method produced the best accuracies, 6.15% higher than the result trained with labeled data only, and 1.3% higher than the conventional methods.

Original languageEnglish
Title of host publicationNeural Information Processing - 24th International Conference, ICONIP 2017, Proceedings
EditorsDongbin Zhao, El-Sayed M. El-Alfy, Derong Liu, Shengli Xie, Yuanqing Li
PublisherSpringer Verlag
Pages854-862
Number of pages9
ISBN (Print)9783319700953
DOIs
Publication statusPublished - 2017 Jan 1
Event24th International Conference on Neural Information Processing, ICONIP 2017 - Guangzhou, China
Duration: 2017 Nov 142017 Nov 18

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10635 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other24th International Conference on Neural Information Processing, ICONIP 2017
CountryChina
CityGuangzhou
Period17/11/1417/11/18

Fingerprint

Semi-supervised Learning
Supervised learning
Labeling
Fusion
Fusion reactions
Dempster-Shafer Theory
Supervised Learning
Labels
High Accuracy
Propagation
Prediction
Experiments
Experiment

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Kim, A., & Cho, S-B. (2017). Dempster-Shafer Fusion of Semi-supervised Learning Methods for Predicting Defaults in Social Lending. In D. Zhao, E-S. M. El-Alfy, D. Liu, S. Xie, & Y. Li (Eds.), Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings (pp. 854-862). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10635 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-70096-0_87
Kim, Aleum ; Cho, Sung-Bae. / Dempster-Shafer Fusion of Semi-supervised Learning Methods for Predicting Defaults in Social Lending. Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings. editor / Dongbin Zhao ; El-Sayed M. El-Alfy ; Derong Liu ; Shengli Xie ; Yuanqing Li. Springer Verlag, 2017. pp. 854-862 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{dbc27fa5735d41f0bd23810645075365,
title = "Dempster-Shafer Fusion of Semi-supervised Learning Methods for Predicting Defaults in Social Lending",
abstract = "In social lending, it is hard to know whether borrowers will repay well or not. Most researchers use supervised learning for default prediction, but labeling data by hand is time-consuming. Moreover, labeling results of semi-supervised learning methods are not the same each other. In this paper, we propose a fusion method of label propagation and transductive SVM based on Dempster-Shafer theory for precisely labeling unlabeled data to improve the performance. We remove few unlabeled data with lower reliabilities in labeling results and fusion of the two results based on Dempster-Shafer theory. We have conducted experiments with supervised learning method trained with labeled unlabeled data. As a result, the proposed method produced the best accuracies, 6.15{\%} higher than the result trained with labeled data only, and 1.3{\%} higher than the conventional methods.",
author = "Aleum Kim and Sung-Bae Cho",
year = "2017",
month = "1",
day = "1",
doi = "10.1007/978-3-319-70096-0_87",
language = "English",
isbn = "9783319700953",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "854--862",
editor = "Dongbin Zhao and El-Alfy, {El-Sayed M.} and Derong Liu and Shengli Xie and Yuanqing Li",
booktitle = "Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings",
address = "Germany",

}

Kim, A & Cho, S-B 2017, Dempster-Shafer Fusion of Semi-supervised Learning Methods for Predicting Defaults in Social Lending. in D Zhao, E-SM El-Alfy, D Liu, S Xie & Y Li (eds), Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10635 LNCS, Springer Verlag, pp. 854-862, 24th International Conference on Neural Information Processing, ICONIP 2017, Guangzhou, China, 17/11/14. https://doi.org/10.1007/978-3-319-70096-0_87

Dempster-Shafer Fusion of Semi-supervised Learning Methods for Predicting Defaults in Social Lending. / Kim, Aleum; Cho, Sung-Bae.

Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings. ed. / Dongbin Zhao; El-Sayed M. El-Alfy; Derong Liu; Shengli Xie; Yuanqing Li. Springer Verlag, 2017. p. 854-862 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10635 LNCS).

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

TY - GEN

T1 - Dempster-Shafer Fusion of Semi-supervised Learning Methods for Predicting Defaults in Social Lending

AU - Kim, Aleum

AU - Cho, Sung-Bae

PY - 2017/1/1

Y1 - 2017/1/1

N2 - In social lending, it is hard to know whether borrowers will repay well or not. Most researchers use supervised learning for default prediction, but labeling data by hand is time-consuming. Moreover, labeling results of semi-supervised learning methods are not the same each other. In this paper, we propose a fusion method of label propagation and transductive SVM based on Dempster-Shafer theory for precisely labeling unlabeled data to improve the performance. We remove few unlabeled data with lower reliabilities in labeling results and fusion of the two results based on Dempster-Shafer theory. We have conducted experiments with supervised learning method trained with labeled unlabeled data. As a result, the proposed method produced the best accuracies, 6.15% higher than the result trained with labeled data only, and 1.3% higher than the conventional methods.

AB - In social lending, it is hard to know whether borrowers will repay well or not. Most researchers use supervised learning for default prediction, but labeling data by hand is time-consuming. Moreover, labeling results of semi-supervised learning methods are not the same each other. In this paper, we propose a fusion method of label propagation and transductive SVM based on Dempster-Shafer theory for precisely labeling unlabeled data to improve the performance. We remove few unlabeled data with lower reliabilities in labeling results and fusion of the two results based on Dempster-Shafer theory. We have conducted experiments with supervised learning method trained with labeled unlabeled data. As a result, the proposed method produced the best accuracies, 6.15% higher than the result trained with labeled data only, and 1.3% higher than the conventional methods.

UR - http://www.scopus.com/inward/record.url?scp=85035142704&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85035142704&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-70096-0_87

DO - 10.1007/978-3-319-70096-0_87

M3 - Conference contribution

AN - SCOPUS:85035142704

SN - 9783319700953

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

SP - 854

EP - 862

BT - Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings

A2 - Zhao, Dongbin

A2 - El-Alfy, El-Sayed M.

A2 - Liu, Derong

A2 - Xie, Shengli

A2 - Li, Yuanqing

PB - Springer Verlag

ER -

Kim A, Cho S-B. Dempster-Shafer Fusion of Semi-supervised Learning Methods for Predicting Defaults in Social Lending. In Zhao D, El-Alfy E-SM, Liu D, Xie S, Li Y, editors, Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings. Springer Verlag. 2017. p. 854-862. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-70096-0_87