Interaction between financial risk measures and machine learning methods

Jun ya Gotoh, Akiko Takeda, Rei Yamamoto

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

The purpose of this article is to review the similarity and difference between financial risk minimization and a class of machine learning methods known as support vector machines, which were independently developed. By recognizing their common features, we can understand them in a unified mathematical framework. On the other hand, by recognizing their difference, we can develop new methods. In particular, employing the coherent measures of risk, we develop a generalized criterion for two-class classification. It includes existing criteria, such as the margin maximization and ν-SVM, as special cases. This extension can also be applied to the other type of machine learning methods such as multi-class classification, regression and outlier detection. Although the new criterion is first formulated as a nonconvex optimization, it results in a convex optimization by employing the nonnegative ℓ 1 -regularization. Numerical examples demonstrate how the developed methods work for bond rating.

Original languageEnglish
Pages (from-to)365-402
Number of pages38
JournalComputational Management Science
Volume11
Issue number4
DOIs
Publication statusPublished - 2014 Sep 27
Externally publishedYes

Fingerprint

Learning systems
Convex optimization
Support vector machines
Risk measures
Learning methods
Machine learning
Financial risk
Interaction
Coherent measures of risk
Regularization
Risk minimization
Bond ratings
Common features
Outlier detection
Support vector machine
Margin

Keywords

  • Coherent measures of risk
  • Conditional value-at-risk (CVaR)
  • Credit rating
  • Mean-absolute semi-deviation (MASD)
  • ν-Support vector machine (ν-SVM)

ASJC Scopus subject areas

  • Management Information Systems
  • Information Systems

Cite this

Interaction between financial risk measures and machine learning methods. / Gotoh, Jun ya; Takeda, Akiko; Yamamoto, Rei.

In: Computational Management Science, Vol. 11, No. 4, 27.09.2014, p. 365-402.

Research output: Contribution to journalArticle

Gotoh, Jun ya ; Takeda, Akiko ; Yamamoto, Rei. / Interaction between financial risk measures and machine learning methods. In: Computational Management Science. 2014 ; Vol. 11, No. 4. pp. 365-402.
@article{d53c7fbdb4a345459339b36c1a1d7ae5,
title = "Interaction between financial risk measures and machine learning methods",
abstract = "The purpose of this article is to review the similarity and difference between financial risk minimization and a class of machine learning methods known as support vector machines, which were independently developed. By recognizing their common features, we can understand them in a unified mathematical framework. On the other hand, by recognizing their difference, we can develop new methods. In particular, employing the coherent measures of risk, we develop a generalized criterion for two-class classification. It includes existing criteria, such as the margin maximization and ν-SVM, as special cases. This extension can also be applied to the other type of machine learning methods such as multi-class classification, regression and outlier detection. Although the new criterion is first formulated as a nonconvex optimization, it results in a convex optimization by employing the nonnegative ℓ 1 -regularization. Numerical examples demonstrate how the developed methods work for bond rating.",
keywords = "Coherent measures of risk, Conditional value-at-risk (CVaR), Credit rating, Mean-absolute semi-deviation (MASD), ν-Support vector machine (ν-SVM)",
author = "Gotoh, {Jun ya} and Akiko Takeda and Rei Yamamoto",
year = "2014",
month = "9",
day = "27",
doi = "10.1007/s10287-013-0175-5",
language = "English",
volume = "11",
pages = "365--402",
journal = "Computational Management Science",
issn = "1619-697X",
publisher = "Springer Verlag",
number = "4",

}

TY - JOUR

T1 - Interaction between financial risk measures and machine learning methods

AU - Gotoh, Jun ya

AU - Takeda, Akiko

AU - Yamamoto, Rei

PY - 2014/9/27

Y1 - 2014/9/27

N2 - The purpose of this article is to review the similarity and difference between financial risk minimization and a class of machine learning methods known as support vector machines, which were independently developed. By recognizing their common features, we can understand them in a unified mathematical framework. On the other hand, by recognizing their difference, we can develop new methods. In particular, employing the coherent measures of risk, we develop a generalized criterion for two-class classification. It includes existing criteria, such as the margin maximization and ν-SVM, as special cases. This extension can also be applied to the other type of machine learning methods such as multi-class classification, regression and outlier detection. Although the new criterion is first formulated as a nonconvex optimization, it results in a convex optimization by employing the nonnegative ℓ 1 -regularization. Numerical examples demonstrate how the developed methods work for bond rating.

AB - The purpose of this article is to review the similarity and difference between financial risk minimization and a class of machine learning methods known as support vector machines, which were independently developed. By recognizing their common features, we can understand them in a unified mathematical framework. On the other hand, by recognizing their difference, we can develop new methods. In particular, employing the coherent measures of risk, we develop a generalized criterion for two-class classification. It includes existing criteria, such as the margin maximization and ν-SVM, as special cases. This extension can also be applied to the other type of machine learning methods such as multi-class classification, regression and outlier detection. Although the new criterion is first formulated as a nonconvex optimization, it results in a convex optimization by employing the nonnegative ℓ 1 -regularization. Numerical examples demonstrate how the developed methods work for bond rating.

KW - Coherent measures of risk

KW - Conditional value-at-risk (CVaR)

KW - Credit rating

KW - Mean-absolute semi-deviation (MASD)

KW - ν-Support vector machine (ν-SVM)

UR - http://www.scopus.com/inward/record.url?scp=84919455614&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84919455614&partnerID=8YFLogxK

U2 - 10.1007/s10287-013-0175-5

DO - 10.1007/s10287-013-0175-5

M3 - Article

VL - 11

SP - 365

EP - 402

JO - Computational Management Science

JF - Computational Management Science

SN - 1619-697X

IS - 4

ER -