On boundary correction in kernel estimation of ROC curves

Warning

This publication doesn't include Faculty of Medicine. It includes Faculty of Science. Official publication website can be found on muni.cz.

Authors	KOLÁČEK Jan KARUNAMUNI Rohana J.
Year of publication	2009
Type	Article in Periodical
Magazine / Source	Austrian Journal of Statistics
MU Faculty or unit	Faculty of Science
Citation
Field	General mathematics
Keywords	kernel estimation; reflection; distribution estimation; ROC curve
Description	The Receiver Operating Characteristic (ROC) curve is a statistical tool for evaluating the accuracy of diagnostics tests. The empirical ROC curve is the most commonly used non-parametric estimator for the ROC curve. To derive a smooth estimate for the ROC curve, we use a kernel smoothing method which has better statistical properties than empirical. By this process we need to estimate a distribution function. It is well known now that kernel distribution estimators are not consistent when estimating a distribution function near the boundary of its support. This problem is known as ``boundary effects". To avoid these difficulties we use the technique, which is a kind of generalized reflection method involving reflecting a transformation of the data. The proposed method generates a class of boundary corrected estimators.
Related projects:	Jaroslav Hájek Center for Theoretical and Applied Statistics