Data envelopment analysis and big data

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989100%3A27510%2F19%3A10239463" target="_blank" >RIV/61989100:27510/19:10239463 - isvavai.cz</a>

Result on the web

<a href="https://www.sciencedirect.com/science/article/pii/S0377221718309123#" target="_blank" >https://www.sciencedirect.com/science/article/pii/S0377221718309123#</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.ejor.2018.10.044" target="_blank" >10.1016/j.ejor.2018.10.044</a>

Result language

angličtina

Original language name

Data envelopment analysis and big data

Original language description

In the traditional data envelopment analysis (DEA) approach for a set of n Decision Making Units (DMUs), a standard DEA model is solved n times, one for each DMU. As the number of DMUs increases, the running-time to solve the standard model sharply rises. In this study, a new framework is proposed to significantly decrease the required DEA calculation time in comparison with the existing methodologies when a large set of DMUs (e.g., 20,000 DMUs or more) is present. The framework includes five steps: (i) selecting a subsample of DMUs using a proposed algorithm, (ii) finding the best-practice DMUs in the selected subsample, (iii) finding the exterior DMUs to the hull of the selected subsample, (iv) identifying the set of all efficient DMUs, and (v) measuring the performance scores of DMUs as those arising from the traditional DEA approach. The variable returns to scale technology is assumed and several simulation experiments are designed to estimate the running-time for applying the proposed method for big data. The obtained results in this study point out that the running-time is decreased up to 99.9% in comparison with the existing techniques. In addition, we illustrate the essential computation time for applying the proposed method as a function of the number of DMUs (cardinality), number of inputs and outputs (dimension), and the proportion of efficient DMUs (density). The methods are also compared on a real data set consisting of 30,099 electric power plants in the United States from 1996 to 2016.

Czech name

—

Czech description

—

Type

J_imp - Article in a specialist periodical, which is included in the Web of Science database

CEP classification

—

OECD FORD branch

10102 - Applied mathematics

Project

<a href="/en/project/GA16-17810S" target="_blank" >GA16-17810S: Selective measures in data envelopment analysis: theory and applications</a><br>

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year

2019

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Name of the periodical

European Journal of Operational Research

ISSN

0377-2217

e-ISSN

—

Volume of the periodical

274

Issue of the periodical within the volume

3

Country of publishing house

NL - THE KINGDOM OF THE NETHERLANDS

Number of pages

8

Pages from-to

1047-1054

UT code for WoS article

000457509200019

EID of the result in the Scopus database

2-s2.0-85056389811