Feature Selection in Data Envelopment Analysis

A Mathematical Optimization approach

Sandra Benítez Peña, Peter Bogetoft, Dolores Romero Morales

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

This paper proposes an integrative approach to feature (input and output) selection in Data Envelopment Analysis (DEA). The DEA model is enriched with zero-one decision variables modelling the selection of features, yielding a Mixed Integer Linear Programming formulation. This single-model approach can handle different objective functions as well as constraints to incorporate desirable properties from the real-world application. Our approach is illustrated on the benchmarking of electricity Distribution System Operators (DSOs). The numerical results highlight the advantages of our single-model approach provide to the user, in terms of making the choice of the number of features, as well as modeling their costs and their nature.
Original languageEnglish
JournalOmega: The International Journal of Management Science
Number of pages24
ISSN0305-0483
DOIs
Publication statusPublished - 30 May 2019

Bibliographical note

Epub ahead of print. Published online: 30. May 2019

Keywords

  • Benchmarking
  • Data envelopment analysis
  • Feature selection
  • Mixed Integer Linear Programming

Cite this

@article{187c9eb44e2740a0b6ca384ee2310748,
title = "Feature Selection in Data Envelopment Analysis: A Mathematical Optimization approach",
abstract = "This paper proposes an integrative approach to feature (input and output) selection in Data Envelopment Analysis (DEA). The DEA model is enriched with zero-one decision variables modelling the selection of features, yielding a Mixed Integer Linear Programming formulation. This single-model approach can handle different objective functions as well as constraints to incorporate desirable properties from the real-world application. Our approach is illustrated on the benchmarking of electricity Distribution System Operators (DSOs). The numerical results highlight the advantages of our single-model approach provide to the user, in terms of making the choice of the number of features, as well as modeling their costs and their nature.",
keywords = "Benchmarking, Data envelopment analysis, Feature selection, Mixed Integer Linear Programming, Benchmarking, Data envelopment analysis, Feature selection, Mixed Integer Linear Programming",
author = "{Ben{\'i}tez Pe{\~n}a}, Sandra and Peter Bogetoft and {Romero Morales}, Dolores",
note = "Epub ahead of print. Published online: 30. May 2019",
year = "2019",
month = "5",
day = "30",
doi = "10.1016/j.omega.2019.05.004",
language = "English",
journal = "Omega: The International Journal of Management Science",
issn = "0305-0483",
publisher = "Elsevier",

}

Feature Selection in Data Envelopment Analysis : A Mathematical Optimization approach. / Benítez Peña, Sandra; Bogetoft, Peter; Romero Morales, Dolores .

In: Omega: The International Journal of Management Science, 30.05.2019.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Feature Selection in Data Envelopment Analysis

T2 - A Mathematical Optimization approach

AU - Benítez Peña, Sandra

AU - Bogetoft, Peter

AU - Romero Morales, Dolores

N1 - Epub ahead of print. Published online: 30. May 2019

PY - 2019/5/30

Y1 - 2019/5/30

N2 - This paper proposes an integrative approach to feature (input and output) selection in Data Envelopment Analysis (DEA). The DEA model is enriched with zero-one decision variables modelling the selection of features, yielding a Mixed Integer Linear Programming formulation. This single-model approach can handle different objective functions as well as constraints to incorporate desirable properties from the real-world application. Our approach is illustrated on the benchmarking of electricity Distribution System Operators (DSOs). The numerical results highlight the advantages of our single-model approach provide to the user, in terms of making the choice of the number of features, as well as modeling their costs and their nature.

AB - This paper proposes an integrative approach to feature (input and output) selection in Data Envelopment Analysis (DEA). The DEA model is enriched with zero-one decision variables modelling the selection of features, yielding a Mixed Integer Linear Programming formulation. This single-model approach can handle different objective functions as well as constraints to incorporate desirable properties from the real-world application. Our approach is illustrated on the benchmarking of electricity Distribution System Operators (DSOs). The numerical results highlight the advantages of our single-model approach provide to the user, in terms of making the choice of the number of features, as well as modeling their costs and their nature.

KW - Benchmarking

KW - Data envelopment analysis

KW - Feature selection

KW - Mixed Integer Linear Programming

KW - Benchmarking

KW - Data envelopment analysis

KW - Feature selection

KW - Mixed Integer Linear Programming

UR - https://sfx-45cbs.hosted.exlibrisgroup.com/45cbs?url_ver=Z39.88-2004&url_ctx_fmt=info:ofi/fmt:kev:mtx:ctx&ctx_enc=info:ofi/enc:UTF-8&ctx_ver=Z39.88-2004&rfr_id=info:sid/sfxit.com:azlist&sfx.ignore_date_threshold=1&rft.object_id=954921388918&rft.object_portfolio_id=&svc.holdings=yes&svc.fulltext=yes

U2 - 10.1016/j.omega.2019.05.004

DO - 10.1016/j.omega.2019.05.004

M3 - Journal article

JO - Omega: The International Journal of Management Science

JF - Omega: The International Journal of Management Science

SN - 0305-0483

ER -