Can AdaBoost.M1 Learn Incrementally? A Comparison to Learn + + Under Different Combination Rules
We had previously introduced Learn + + , inspired in part by the ensemble based AdaBoost algorithm, for incrementally learning from new data, including new concept classes, without forgetting what had been previously learned. In this effort, we compare the incremental learning performance of Learn + + and AdaBoost under several combination schemes, including their native, weighted majority voting. We show on several databases that changing AdaBoost’s distribution update rule from hypothesis based update to ensemble based update allows significantly more efficient incremental learning ability, regardless of the combination rule used to combine the classifiers.
KeywordsGeneralization Performance Incremental Learning Combination Rule Weak Classifier Narrow Confidence Interval
Unable to display preview. Download preview PDF.
- 6.Polikar, R., Byorick, J., Krause, S., Marino, A., Moreton, M.: Learn++: A classifier independent incremental learning algorithm for supervised neural networks. In: Proc. of Int. Joint Conference on Neural Networks (IJCNN 2002), May 12-17, 2002, vol. 2, pp. 1742–1747. Honolulu, HI (2002)Google Scholar
- 10.Blake, C.L., Merz, C.J.: Univ. of California, Irvine, Repository of Machine Learning Databases at Irvine, CAGoogle Scholar