Nonparametric predictive inference with parametric copulas for combining bivariate diagnostic tests
Abstract
Measuring the accuracy of diagnostic tests is crucial in many application areas including medicine, machine learning and credit scoring. The receiver operating characteristic (ROC) curve is a useful tool to assess the ability of a diagnostic test to discriminate among two classes or groups. In practice, multiple diagnostic tests or biomarkers may be combined to improve diagnostic accuracy, e.g. by maximizing the area under the ROC curve. In this paper we present Nonparametric Predictive Inference (NPI) for best linear combination of two biomarkers, where the dependence of the two biomarkers is modelled using parametric copulas. NPI is a frequentist statistical method that is explicitly aimed at using few modelling assumptions, enabled through the use of lower and upper probabilities to quantify uncertainty. The combination of NPI for the individual biomarkers, combined with a basic parametric copula to take dependence into account, has good robustness properties and leads to quite straightforward computation. We briefly comment on the results of a simulation study to investigate the performance of the proposed method in comparison to the empirical method. An example with data from the literature is provided to illustrate the proposed method, and related research problems are briefly discussed.References
T. Augustin and F.P.A. Coolen. Nonparametric predictive inference and interval probability. Journal of Statistical Planning and Inference, vol. 124, pp. 251–272, 2004.
T. Augustin, F.P.A. Coolen, G. de Cooman and M.C.M. Troffaes (Eds). Introduction to Imprecise Probabilities. Chichester: Wiley, 2014.
A. Bansal and M. Sullivan Pepe. When does combining markers improve classification performance and what are implications for practice? Statistics in Medicine, vol. 32, pp. 1877–1892, 2013.
F.P.A. Coolen. On nonparametric predictive inference and objective Bayesianism. Journal of Logic, Language and Information, vol. 15, pp. 21–47, 2006.
T. Coolen-Maturi. Three-group ROC predictive analysis for ordinal outcomes. Communications in Statistics - Theory and Methods, vol. 46, pp. 9476-9493, 2017.
T. Coolen-Maturi, F.P.A. Coolen and N. Muhammad. Predictive inference for bivariate data: combining nonparametric predictive inference for marginals with an estimated copula. Journal of Statistical Theory and Practice, vol. 10, pp. 515–538, 2016.
T. Coolen-Maturi, P. Coolen-Schrijner and F.P.A. Coolen. Nonparametric predictive inference for binary diagnostic tests. Journal of Statistical Theory and Practice, vol. 6, pp. 665–680, 2012.
T. Coolen-Maturi, P. Coolen-Schrijner and F.P.A. Coolen. Nonparametric predictive inference for diagnostic accuracy. Journal of Statistical Planning and Inference, vol. 142, pp. 1141–1150, 2012.
T. Coolen-Maturi, F.F. Elkhafifi and F.P.A. Coolen. Three-group ROC analysis: A nonparametric predictive approach. Computational Statistics & Data Analysis, vol. 78, pp. 69–81, 2014.
B. De Finetti. Theory of Probability. London: Wiley, 1974.
F.F. Elkhafifi and F.P.A. Coolen. Nonparametric predictive inference for accuracy of ordinal diagnostic tests. Journal of Statistical Theory and Practice, vol. 6, pp. 681–697, 2012.
B.M. Hill. Posterior distribution of percentiles: Bayes’ theorem for sampling from a population. Journal of the American Statistical Association, vol. 63, pp. 677–691, 1968.
L. Kang, A. Liu and L. Tian. Linear combination methods to improve diagnostic/prognostic accuracy on future observations. Statistical Methods in Medical Research, vol. 25, pp. 1359–1380, 2016.
J.F. Lawless and M. Fredette. Frequentist prediction intervals and predictive distributions. Biometrika, vol. 92, pp. 529–542, 2005.
C. Liu, A. Liu and S. Halabi. A min–max combination of biomarkers to improve diagnostic accuracy. Statistics in Medicine, vol. 30, pp. 2005–2014, 2011.
N. Muhammad. Predictive Inference with Copulas for Bivariate Data. PhD thesis, Durham University, UK (available from www.npistatistics.com), 2016.
J.Q. Su and J.S. Liu. Linear combinations of multiple diagnostic markers. Journal of the American Statistical Association, vol. 88, pp. 1350–1355, 1993.
M. Sullivan Pepe. The Statistical Evaluation of Medical Tests for Classification and Prediction. Oxford: Oxford University Press, 2003.
M. Sullivan Pepe and M.L. Thompson. Combining diagnostic test results to increase accuracy. Biostatistics, vol. 1, pp. 123–140, 2000.
S. Wieand, M.H. Gail, B.R. James and K.L. James. A family of nonparametric statistics for comparing diagnostic markers with paired or unpaired data. Biometrika, vol. 76, pp. 585–592, 1989.
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
