A unifying approach on bias and variance analysis for classification
This work provides a more coherent theoretical understanding of bias and variance for researchers and practitioners working with classification models, addressing the problem of disparate definitions in the field.
This paper unifies the Tumer & Ghosh and James frameworks for bias and variance analysis in classification. It establishes closed-form relationships between bias and variance defined for 0/1 loss and those for squared error loss of boundary distributions, enhancing understanding of classification performance.
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.