Ensembling Imbalanced-Spatial-Structured Support Vector Machine

Authors

Xin Liu, University of Waterloo
Grace Y. Yi, University of Waterloo
Glenn Bauman, The University of Western OntarioFollow
Wenqing He, The University of Western Ontario

Document Type

Article

Publication Date

1-1-2021

Journal

Econometrics and Statistics

Volume

17

First Page

145

Last Page

155

URL with Digital Object Identifier

10.1016/j.ecosta.2020.02.003

Abstract

The support vector machine (SVM) and its extensions have been widely used in various areas. However, these methods cannot effectively handle imbalanced data with spatial association. The ensembling imbalanced-spatial-structured support vector machine (EISS-SVM) method is proposed to handle such data. Not only the proposed method accommodates the relationship between the response and predictors, but also accounts for the spatial correlation existing in data which may be imbalanced. The EISS-SVM classifier embraces the usual SVM as a special case. Numerical studies show satisfactory performance of the proposed method, and the analysis results are reported for the application of the proposed method to handling the imaging data from an ongoing prostate cancer research conducted in Canada.