Title
Cost-based feature selection for Support Vector Machines: An application in credit scoring.
Abstract
•We present two mixed-integer problems for cost-based attribute selection.•The attribute acquisition costs are incorporated explicitly in our framework.•The attribute acquisition costs are estimated for a credit scoring project.•Our method leads to a ten-fold reduction in costs in a real-life lending dataset.
Year
DOI
Venue
2017
10.1016/j.ejor.2017.02.037
European Journal of Operational Research
Keywords
Field
DocType
Analytics,Feature selection,Support Vector Machines,Mixed-integer programming,Credit scoring
Data mining,Feature selection,Computer science,Multivariate statistics,Support vector machine,Integer programming,Linear programming,Artificial intelligence,Analytics,Machine learning
Journal
Volume
Issue
ISSN
261
2
0377-2217
Citations 
PageRank 
References 
19
0.60
18
Authors
3
Name
Order
Citations
PageRank
Sebastián Maldonado150832.45
Juan F. Pérez210611.80
Cristián Bravo312410.63