Distance Weighted Discrimination and Second Order Cone Programming

Abstract

This vignette demonstrates the utility and flexibility of the R-package dwd in conducting classification and optimization problems. Distance Weighted Discrimination (DWD) is a recently developed powerful classification method which was originally motivated for solving the High Dimensional Low Sample Size (HDLSS) examples (Marron et al. (2007)), but can be… (More)

Topics

  • Presentations referencing similar topics