Novel Approach to Gentle AdaBoost Algorithm with Linear Weak Classifiers

Abstract

This paper presents the problem of calculating the value of the scoring function for weak classifiers operating in the sequential structure. An example of such a structure is Gentle AdaBoost algorithm whose modification we propose in this work. In the proposed approach the distance of the object from the decision boundary is scaled in decision regions defined by the weak classifier at first and later transformed by the log-normal function. The described algorithm was tested on sixth public available data sets and compared with Gentle AdaBoost algorithm.