AdaBoost for Feature Selection, Classification and Its Relation with SVM*, A Review

In order to clarify the role of AdaBoost algorithm for feature selection, classifier learning and its relation with SVM, this paper provided a brief introduction to the AdaBoost which is used for producing a strong classifier out of weak learners firstly. The original adaptive boosting algorithm and its application in face detection and facial expression recognition are reviewed. In pattern classification domain, support vector machine has been widely used and shows promising performance. However, it is expensive in terms of time-consuming. A sort of cascaded support vector machines architecture is capable of improving the classification accuracy based on AdaBoost boosting algorithm, namely, AdaboostSVM. It applied boosting algorithm to feature selection and classifier learning for support vector machine classification and it has achieved approved performance through some researcher's pioneering work. (C) 2012 Published by Elsevier B.V. Selection and/or peer-review under responsibility of Garry Lee