A new ensemble-based cascaded framework for multiclass training with simple weak learners

Abstract

We present a novel approach to multiclass learning using an ensemble-based cascaded learning framework. By implementing a multiclass cascaded classifier with AdaBoost, we show how detection runtimes are accelerated since only a subset of the ensemble is executed, thus making the classifiers suitable for computer vision applications. We also propose a new multiclass weak learner and demonstrate the framework's ability to achieve arbitrarily low training errors in conjunction with it. We tested our algorithm against AdaBoost.OC, ECC and M2 multiclass learning methods, on seven benchmark UCI datasets. In our experiments, we found that our framework achieves higher accuracy on five out of seven datasets and displays faster runtime efficiency in all cases. © 2011 Springer-Verlag.