On the use of different base classifiers in multiclass problems

• Published in: Progress in artificial intelligence Vol. 6; no. 4; pp. 315 - 323
• Main Authors: Morán-Fernández, L., Bolón-Canedo, V., Alonso-Betanzos, A.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01-12-2017; Springer Nature B.V
• Subjects: Adequacy; Artificial Intelligence; Classifiers; Computational Intelligence; Computer Imaging; Computer Science; Control; Data Mining and Knowledge Discovery; Mechatronics; Natural Language Processing (NLP); Pattern Recognition and Graphics; Regular Paper; Robotics; Vision; One-versus-one; Base classifier; Multiclass classification; Data complexity measures; Binarization techniques
• ISSN: 2192-6352; 2192-6360
• DOI: 10.1007/s13748-017-0126-4

Classification problems with more than two classes can be handled in different ways. The most used approach is the one which transforms the original multiclass problem into a series of binary subproblems which are solved individually. In this approach, should the same base classifier be used on all binary subproblems? Or should these subproblems be tuned independently? Trying to answer this question, in this paper we propose a method to select a different base classifier in each subproblem—following the one-versus-one strategy—making use of data complexity measures. The experimental results on 17 real-world datasets corroborate the adequacy of the method.