Vote-boosting ensembles. (November 2018)

• Record Type: Journal Article
• Authors: Sabzevari, Maryam; Martínez-Muñoz, Gonzalo; Suárez, Alberto
• Is Part Of: Pattern recognition. Volume 83(2018:Nov.)
• Page Start: 119
• Page End: 133
• Publication Date: 2018-11
• Subjects: Ensemble learning -- Boosting -- Uncertainty-based emphasis -- Robust classification; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2018.05.022
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: A boosting algorithm with a new emphasis function is proposed. The instance weights are determined in terms of the degree of agreement or disagreement among the individual ensemble predictions. The optimal type of emphasis (either on instances for which there is agreement or disagreement) can be empirically determined using cross validation. Vote-boosting can be used to build ensembles that are both accurate and robust to class-label noise. Abstract: Vote-boosting is a sequential ensemble learning method in which the individual classifiers are built on different weighted versions of the training data. To build a new classifier, the weight of each training instance is determined in terms of the degree of disagreement among the current ensemble predictions for that instance. For low class-label noise levels, especially when simple base learners are used, emphasis should be made on instances for which the disagreement rate is high. When more flexible classifiers are used and as the noise level increases, the emphasis on these uncertain instances should be reduced. In fact, at sufficiently high levels of class-label noise, the focus should be on instances on which the ensemble classifiers agree. The optimal type of emphasis can be automatically determined using cross-validation. An extensive empirical analysis using the beta distribution as emphasis function illustrates that vote-boosting is an effective method to generate ensembles that are both accurate and robust.