Information Theoretic Regularization for Semi-Supervised Boosting

TitleInformation Theoretic Regularization for Semi-Supervised Boosting
Publication TypeConference Paper
Year of Publication2009
AuthorsLei Zheng, Yan Liu, Shaojun Wang
Conference NameKnowledge Discovery and Data Mining - KDD2009
Conference LocationParis, France
Abstract

We present novel semi-supervised boosting algorithms that incrementally build linear combinations of weak classifiers through generic functional gradient descent using both labeled and unlabeled training data. Our approach is based on extending information regularization framework to boosting,bearing loss functions that combine log loss on labeled data with the information-theoretic measures to encode unlabeled data. Even though the information-theoretic regularization terms make the optimization non-convex, we propose simple sequential gradient descent optimization algorithms, and obtain impressively improved results on synthetic, benchmark and real world tasks over supervised boosting algorithms which use the labeled data alone and a state-of-the-art semi-supervised boosting algorithm

Full Text

L. Zheng, S. Wang, Y. Liu and C. Lee 'Information theoretic regularization for semi-supervised boosting', Knowledge Discovery and Data Mining - KDD2009, Paris, June 28 - July 1, 2009
year: 2009
venue name: Knowledge Discovery and Data Mining - KDD2009
hasURL: http://knoesis.wright.edu/library/download/kdd09.pdf

Related Files: