A0760
Title: Semi-parametric confidence bands for the ROC curve based on unions of transformed elliptical regions
Authors: Andrew Spieker - Vanderbilt University Medical Center (United States) [presenting]
Abstract: Receiver operating characteristic (ROC) analysis is a helpful tool to evaluate the capacity of a measure to distinguish between states. Assumptions regarding the structure of the ROC curve can improve the efficiency of estimation for the ROC curve itself or its associated area under the curve (AUC). Many are familiar with nonparametric methods for generating an ROC curve; fewer are familiar with parametric and semi-parametric methods. Because the ROC curve is a two-dimensional projection of a fundamentally three-dimensional object, the challenges of characterizing uncertainty in the ROC curve or its associated AUC are often overlooked. The focus is on methods to characterize uncertainty based on a unions-of-transformed-ellipses approach.
