Classifier Combination as a Tomographic Process

Abstract

A mathematical analogy between the process of multiple expert fusion and the tomographic reconstruction of Radon integral data is outlined for the specific instance of the combination of classifiers containing discrete data sets. Within this metaphor all conventional methods of classifier combination come, to a greater or lesser degree, to resemble the unfiltered back-projection of the constituent classifiers’ probability density functions: an implicit attempt to reconstruct the PDF of the composite pattern space. In these probabilistic terms, the combination of classifiers with identical feature-sets correspondingly constitutes an attempt at morphological manipulation of the composite pattern-space PDF. A consideration of the separate benefits of combination along these dualistic lines eventually leads to an optimal strategy for classifier combination under arbitrary conditions.