On the Convergence of Leveraging

We give an unified convergence analysis of ensemble learning methods including e.g. AdaBoost, Logistic Regression and the Least-SquareBoost algorithm for regression. These methods have in common that they iteratively call a base learning algorithm which returns hypotheses that are then linearly combined. We show that these methods are related to the Gauss-Southwell method known from numerical optimization and state non-asymptotical convergence results for all these methods. Our analysis includes

• Type: Conference
• Year: 2001
• Where: NIPS
• Authors: Gunnar Rätsch, Sebastian Mika, Manfred K. Warmuth
• Keywords: Base Learning Algorithm, Ensemble Learning Methods, NIPS 2001, NIPS 2007, Unified Convergence Analysis