A comparative analysis of instance-based penalization techniques for classification
Several instance-based large-margin classifiers have recently been put forward in the literature: Support Hyperplanes, Nearest Convex Hull classifier and Soft Nearest Neighbor. We examine those techniques from a common fit-versuscomplexity framework and study the links between them. Finally, we compare the performance of these techniques vis-a-vis each other and other standard classification methods.
|Organisation||Department of Econometrics|
Nalbantov, G.I, Groenen, P.J.F, & Smirnov, E. (2012). A comparative analysis of instance-based penalization techniques for classification. doi:10.1007/978-1-4614-1903-7