Boosting through optimization of margin distributions

Date

2010

Authors

Shen, Chunhua
Li, Hanxi

Journal Title

Journal ISSN

Volume Title

Publisher

Institute of Electrical and Electronics Engineers (IEEE Inc)

Abstract

Boosting has been of great interest recently in the machine learning community because of the impressive performance for classifi- cation and regression problems. The success of boosting algorithms may be interpreted in terms of the margin theory. Recently, it has been shown that generalization error of classifiers can be obtained by explicitly taking the margin distribution of the training data into account. Most of the current boosting algorithms in practice usually optimize a convex loss function and do not make use of the margin distribution. In this brief, we design a new boosting algorithm, termed margin-distribution boosting (MDBoost), which directly maximizes the average margin and minimizes the margin variance at the same time. This way the margin distribution is optimized. A totally corrective optimization algorithm based on column generation is proposed to implement MDBoost. Experiments on various data sets show that MDBoost outperforms AdaBoost and LPBoost in most cases.

Description

Keywords

Keywords: AdaBoost; Boosting algorithm; Column generation; Data sets; Generalization Error; Loss functions; Machine learning communities; Margin theory; Optimization algorithms; Regression problem; Training data; Algorithms; Optimization; Adaptive boosting; algorit AdaBoost; Boosting; Column generation; Margin distribution

Citation

Source

IEEE Transactions on Neural Networks

Type

Journal article

Book Title

Entity type

Access Statement

License Rights

Restricted until

2037-12-31