Information theoretic regularization for semi-supervised boosting

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 semisupervised boosting algorithm. Categories and Subject Descriptors I.2 [Artificial Intelligence]: Learning General Terms Algorithms, Experimentation, Performance Keywords Ensemble method, semi-supervised learning

• Type: Conference
• Year: 2009
• Where: KDD
• Authors: Lei Zheng, Shaojun Wang, Yan Liu, Chi-Hoon Lee
• Keywords: Data Mining, Descent Optimization Algorithms, Information-theoretic Regularization Terms, KDD 2009, Semi-supervised Boosting Algorithms