Statistics and Risk Management Seminar - 9/12/2015

Wednesday, 9 December 2015, 3:00 p.m.

Lecturer: Vanda Inácio, Assistant Professor, Department of Statistics, Pontificia Universidad Católica de Chile

Title: "Bayesian bootstrap inference for the ROC surface"

Local: Room 1.3, Edifício VII

Faculdade de Ciências e Tecnologia, Quinta da Torre, Caparica

Abstract: The receiver operating characteristic (ROC) surface is a popular tool for evaluating the accuracy of medical diagnostic tests that classify individuals into one of three ordered classes. We propose a fully nonparametric method based on the Bayesian bootstrap for conducting inferences for the ROC surface and its functionals, such as the volume under the surface. The proposed estimator is based on a simple, yet interesting, representation of the ROC surface in terms of placement variables. A simulation study, comparing our estimator with its main competitors, is conducted, and data concerning the diagnosis of neurological impairment in HIV patients is analyzed.