RT Journal Article
T1 Robust consistent estimators for ROC curves with covariates
A1 Bianco, Ana Maria
A1 Boente Boente, Graciela
A1 González Manteiga, Wenceslao
K1 Covariates
K1 Robustness
K1 ROC curves
K1 Parametric regression
AB The Receiver Operating Characteristic (ROC) curve is a usefultool to measure the classification capability of a continuous variable toassess the accuracy of a medical test that distinguishes between two conditions.Sometimes, covariates related to the diagnostic variable may increasethe discriminating power of the ROC curve. Due to the lack of stability ofclassical ROC curves estimators to outliers, we introduce a procedure toobtain robust estimators in presence of covariates. The considered proposalfocusses on a semiparametric approach which robustly fits a location-scaleregression model to the diagnostic variable and considers robust adaptiveempirical estimators of the regression residuals. The uniform consistencyof the proposal is derived under mild assumptions. A Monte Carlo study iscarried out to compare the performance of the robust proposed estimatorswith the classical ones both, in clean and contaminated samples. A realdata set is also analysed.
PB EUCLID
SN 1935-7524
YR 2022
FD 2022
LK http://hdl.handle.net/10347/33474
UL http://hdl.handle.net/10347/33474
LA eng
NO Bianco, A. M., Boente, G., & González–manteiga, W. (2022). Robust consistent estimators for ROC curves with covariates. Electronic Journal of Statistics, 16(2), 4133-4161.
NO This research was partially supported by Grants PICT 2018-00740 from ANPCYT and 20020170100022BA from the Universidad de Buenos Aires, Argentina and also by the Spanish Projects MTM2016-76969P and PID2020-116587GB-I00 from the from the Ministry of Science and Innovation (MCIN/ AEI/FEDER, UE), Spain.
DS Minerva
RD 29 abr 2026