Bayesian bootstrap estimation of ROC curve

被引:40
作者
Gu, Jiezhun [1 ]
Ghosal, Subhashis [2 ]
Roy, Anindya [3 ]
机构
[1] Duke Univ, Med Ctr, Duke Clin Res Inst, Durham, NC 27715 USA
[2] N Carolina State Univ, Dept Stat, Raleigh, NC 27695 USA
[3] Univ Maryland, Dept Math & Stat, Baltimore, MD 21201 USA
关键词
area under the curve (AUC); Bayesian bootstrap; integrated absolute error; ROC Curve; testing binormality;
D O I
10.1002/sim.3366
中图分类号
Q [生物科学];
学科分类号
07 ; 0710 ; 09 ;
摘要
Receiver operating characteristic (ROC) Curve is widely applied in measuring discriminatory ability of diagnostic or prognostic tests. This makes the ROC analysis one of the most active research areas in medical statistics. Many parametric and semiparametric estimation methods have been proposed for estimating the ROC curve and its functionals. In this paper, we propose the Bayesian bootstrap (BB), a fully nonparametric estimation method, for the ROC curve and its functionals, such as the area under the curve (AUC). The BB method offers a bandwidth-free smoothing approach to the empirical estimate, and gives credible bounds. The accuracy of the estimate of the ROC curve in the simulation studies is examined by the integrated absolute error. In comparison with other existing curve estimation methods, the BB method performs well in terms of accuracy, robustness and simplicity. We also propose a procedure based on the BB approach to test the binormality assumption. Copyright (C) 2008 John Wiley & Sons, Ltd.
引用
收藏
页码:5407 / 5420
页数:14
相关论文
共 35 条
[1]   AREA ABOVE ORDINAL DOMINANCE GRAPH AND AREA BELOW RECEIVER OPERATING CHARACTERISTIC GRAPH [J].
BAMBER, D .
JOURNAL OF MATHEMATICAL PSYCHOLOGY, 1975, 12 (04) :387-415
[2]  
BOZDOGAN H, 1986, COMPUT STAT Q, V3, P127
[3]   ESTIMATING PR(X-LESS-THAN-Y) IN CATEGORIZED DATA USING ROC ANALYSIS [J].
BROWNIE, C .
BIOMETRICS, 1988, 44 (02) :615-621
[4]   Semi-parametric estimation of the binormal ROC curve for a continuous diagnostic test [J].
Cai, TX ;
Moskowitz, CS .
BIOSTATISTICS, 2004, 5 (04) :573-586
[5]   COMPARING THE AREAS UNDER 2 OR MORE CORRELATED RECEIVER OPERATING CHARACTERISTIC CURVES - A NONPARAMETRIC APPROACH [J].
DELONG, ER ;
DELONG, DM ;
CLARKEPEARSON, DI .
BIOMETRICS, 1988, 44 (03) :837-845
[6]   Proper receiver operating characteristic analysis: The bigamma model [J].
Dorfman, DD ;
Berbaum, KS ;
Metz, CE ;
Lenth, RV ;
Hanley, JA ;
AbuDagga, H .
ACADEMIC RADIOLOGY, 1997, 4 (02) :138-149
[7]   MAXIMUM LIKELIHOOD ESTIMATION OF PARAMETERS OF SIGNAL DETECTION THEORY - A DIRECT SOLUTION [J].
DORFMAN, DD ;
ALF, E .
PSYCHOMETRIKA, 1968, 33 (01) :117-&
[8]   1977 RIETZ LECTURE - BOOTSTRAP METHODS - ANOTHER LOOK AT THE JACKKNIFE [J].
EFRON, B .
ANNALS OF STATISTICS, 1979, 7 (01) :1-26
[9]   RECEIVER OPERATOR CHARACTERISTIC (ROC) CURVES AND NONNORMAL DATA - AN EMPIRICAL-STUDY [J].
GODDARD, MJ ;
HINBERG, I .
STATISTICS IN MEDICINE, 1990, 9 (03) :325-337
[10]  
GREEN DM, SIGNAL DETECTION THE