Classification of countries' progress toward a knowledge economy based on machine learning classification techniques

Áreas de investigación:

• Methodology - Machine learning

Año:

2015

Tipo de publicación:

Artículo

Palabras clave:

Decision support systems, Machine learning, Knowledge economy, Hierarchical clustering, Ordinal classification

Autores:

• de la Paz Marín, Mónica
• Gutiérrez, Pedro Antonio
• Hervás-Martínez, César

Journal:

Expert Systems with Applications

Volumen:

42

Número:

1

Páginas:

562-572

Mes:

January

ISSN:

0957-4174

BibTex:

@article{ESWA2014Monica, author = "M{\'o}nica de la Paz Mar{\'i}n and Pedro Antonio Guti{\'e}rrez and C{\'e}sar Herv{\'a}s-Mart{\'i}nez", abstract = "Knowledge is a key factor of competitive advantages in the current economic crisis and uncertain environment. There are a number of indicators to measure knowledge advances, however, the benefits for stakeholders and policy makers are limited because of a lack of classification models. This paper introduces an approach to classify 54 countries (in 2007--2009) according to their progress toward a knowledge economy (KE). To achieve this, the aims of this paper are twofold: first, to find clusters of countries at a similar stage of development toward KE to test if they are meaningful; hence, it will be possible to order the clusters from early KEs (last cluster) to advanced KEs (first cluster). Second, having obtained these clusters, it is possible to build various models to detect the advancement of countries toward KE from one year to another due to its classification. Then, three ordinal classifiers from the machine-learning field were compared in order to select the classifier that performs the best and to confirm the ordinal description of the clusters. Finally, an ordinal model based on the Support Vector Ordinal Regression with Implicit Constraints was selected because of its ability to classify the patterns into the clusters, confirming the appropriateness of the clusters and their ordinal nature. The proposed ordinal classifier could be used for monitoring the progress or stage of transition to KE and for analysing whether a country changes clusters, entering one that performs better or worse.", awards = "JCR(2015): 2.981 Position: 6/82 (Q1) Category: OPERATIONS RESEARCH {\&}amp; MANAGEMENT SCIENCE", comments = "JCR(2015): 2.981 Position: 6/82 (Q1) Category: OPERATIONS RESEARCH {\&}amp; MANAGEMENT SCIENCE", doi = "10.1016/j.eswa.2014.08.008", issn = "0957-4174", journal = "Expert Systems with Applications", keywords = "Decision support systems, Machine learning, Knowledge economy, Hierarchical clustering, Ordinal classification", month = "January", note = "JCR(2015): 2.981 Position: 6/82 (Q1) Category: OPERATIONS RESEARCH {\&}amp; MANAGEMENT SCIENCE", number = "1", pages = " 562--572", title = "{C}lassification of countries{\&}#039; progress toward a knowledge economy based on machine learning classification techniques", url = "http://dx.doi.org/10.1016/j.eswa.2014.08.008", volume = "42", year = "2015", }

Nota:

JCR(2015): 2.981 Position: 6/82 (Q1) Category: OPERATIONS RESEARCH & MANAGEMENT SCIENCE

Abstract:

Knowledge is a key factor of competitive advantages in the current economic crisis and uncertain environment. There are a number of indicators to measure knowledge advances, however, the benefits for stakeholders and policy makers are limited because of a lack of classification models. This paper introduces an approach to classify 54 countries (in 2007--2009) according to their progress toward a knowledge economy (KE). To achieve this, the aims of this paper are twofold: first, to find clusters of countries at a similar stage of development toward KE to test if they are meaningful; hence, it will be possible to order the clusters from early KEs (last cluster) to advanced KEs (first cluster). Second, having obtained these clusters, it is possible to build various models to detect the advancement of countries toward KE from one year to another due to its classification. Then, three ordinal classifiers from the machine-learning field were compared in order to select the classifier that performs the best and to confirm the ordinal description of the clusters. Finally, an ordinal model based on the Support Vector Ordinal Regression with Implicit Constraints was selected because of its ability to classify the patterns into the clusters, confirming the appropriateness of the clusters and their ordinal nature. The proposed ordinal classifier could be used for monitoring the progress or stage of transition to KE and for analysing whether a country changes clusters, entering one that performs better or worse.

Comentarios:

JCR(2015): 2.981 Position: 6/82 (Q1) Category: OPERATIONS RESEARCH & MANAGEMENT SCIENCE