Can I read Variable Importance in Imbalanced Datasets on EtoBox?

Variable Importance in Imbalanced Datasets by Vedic Krishna is a document available to read on EtoBox.

What is Variable Importance in Imbalanced Datasets about?

This document presents two new techniques for variable importance analysis (VIA) in imbalanced datasets for multiclass classification problems: mh-χ2 and a parametric method based on global sensitivity analysis (GSA). The mh-χ2 technique employs a multivariate continuous response framework to deal with multiclass classification, capturing dissimilarities in misclassification errors before and after permuting input variables. The GSA solution quantifies output variability due to input variable variance using

Author

Vedic Krishna

Language

EN