Boosting Trees for Cost-Sensitive Classi cations

Abstract

This paper explores two boosting techniques for cost-sensitive tree classi cations in the situation where misclassi cation costs change very often. Ideally, one would like to have only one induction, and use the induced model for di erent misclassi cation costs. Thus, it demands robustness of the induced model against cost changes. Combining multiple trees… (More)

1 Figure or Table

Topics

  • Presentations referencing similar topics