Multi-class cost-sensitive boosting with p-norm loss functions

We propose a family of novel cost-sensitive boosting methods for multi-class classification by applying the theory of gradient boosting to p-norm based cost functionals. We establish theoretical guarantees including proof of convergence and convergence rates for the proposed methods. Our theoretical treatment provides interpretations for some of the existing algorithms in terms of the proposed family, including a generalization of the costing algorithm [16], DSE and GBSE-t [1], and the Average Cost method [8]. We also experimentally evaluate the performance of our new algorithms against existing methods of cost-sensitive boosting, including AdaCost [5], CSB2 [13], and AdaBoost.M2 [6] with costsensitive weight initialization. We show that our proposed scheme generally achieves superior results in terms of cost minimization and, with the use of higher order p-norm loss in certain cases, consistently outperforms the compararison methods, thus establishing its empirical advantage.

• Type: Conference
• Year: 2008
• Where: KDD
• Authors: Aurelie C. Lozano, Naoki Abe
• Keywords: Average Cost Method, Cost-sensitive Boosting Methods, Data Mining, Gradient Boosting, KDD 2008