Classifier Combination

Abstract

In this chapter we give a survey of the combination of classifiers. Section 14.1 to 14.3 briefly describe basic principles of machine learning and the problem of classifier construction. Section 14.4 reviews several approaches to generate different classifiers as well as established methods to combine different classi- fiers. Then, from Section 14.5, we introduce our novel approach to assess the appropriateness of different classifiers based on their characteristics for each test point individually.