LogitROC: An R Package for Making Inferences on ROC Curves and Surfaces Using Nonparametric and Semiparametric Approaches

• Activity Number: 167
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Section for Statistical Programmers and Analysts
• Author(s): Dong Zhang*+
• Companies: Bloomsburg University
• Keywords: Density ratio model ; Empirical likelihood ; Receiver operating characteristic (ROC) ; Area under the ROC curve (AUC)

Abstract

LogitROC is an R package for simulation, visualization and estimation of receiver operating characteristic (ROC) curves and surfaces using both nonparametric and semiparametric approaches. Empirical likelihood estimators and semiparametric estimators considering the density ratio model of ROC curves and surfaces and comparison between correlated ROC curves are generated. Plots of estimated ROC curves or surfaces, and estimated distribution functions are produced for different parameter values. In addition, empirical likelihood and semiparametric confidence intervals are proposed. Especially, on the Windows system, there is an interactive feature for the user to find the specific sensitivity and specificity by clicking the ROC curves or surfaces.