2013 | OriginalPaper | Buchkapitel
Smoothed Emphasis for Boosting Ensembles
verfasst von : Anas Ahachad, Adil Omari, Aníbal R. Figueiras-Vidal
Erschienen in: Advances in Computational Intelligence
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Real AdaBoost ensembles have exceptional capabilities for successfully solving classification problems. This characteristic comes from progressively constructing learners paying more attention to samples that are difficult to be classified.
However, the corresponding emphasis can be excessive. In particular, when the problem to solve is asymmetric or includes imbalanced outliers, even the previously proposed modifications of the basic algorithm are not as effective as desired.
In this paper, we introduce a simple modification which uses the neighborhood concept to reduce the above drawbacks. Experimental results confirm the potential of the proposed scheme.
The main conclusions of our work and some suggestions for further research along this line close the paper.