Confidence estimation using the incremental learning algorithm, learn++

Abstract

Pattern recognition problems span a broad range of applications, where each application has its own tolerance on classification error. The varying levels of risk associated with many pattern recognition applications indicate the need for an algorithm with the ability to measure its own confidence. In this work, the supervised incremental learning algorithm Learn++ [1], which exploits the synergistic power of an ensemble of classifiers, is further developed to add the capability of assessing its own confidence using a weighted exponential majority voting technique. © Springer-Verlag Berlin Heidelberg 2003.