A unifying approach on bias and variance analysis for classification

TL;DR

This paper unifies different bias and variance analysis frameworks for classification, linking regression-based concepts to classification-specific models, and demonstrates their practical use through case studies.

Contribution

It provides a unified theoretical framework connecting Tumer & Ghosh and James approaches to bias and variance in classification.

Findings

Unified bias and variance models for classification

Closed-form relationships between different B&V frameworks

Enhanced understanding of classification performance

Abstract

Standard bias and variance (B&V) terminologies were originally defined for the regression setting and their extensions to classification have led to several different models / definitions in the literature. In this paper, we aim to provide the link between the commonly used frameworks of Tumer & Ghosh (T&G) and James. By unifying the two approaches, we relate the B&V defined for the 0/1 loss to the standard B&V of the boundary distributions given for the squared error loss. The closed form relationships provide a deeper understanding of classification performance, and their use is demonstrated in two case studies.