Academic Journal
A jackknife approach to estimate the prediction uncertainty from binary classifiers under right-censoring.
| Τίτλος: | A jackknife approach to estimate the prediction uncertainty from binary classifiers under right-censoring. |
|---|---|
| Συγγραφείς: | Jahn-Eimermacher A; Department of Mathematics and Natural Sciences, University of Applied Sciences, Darmstadt, Germany.; European University of Technology, European Union, Darmstadt, Germany., Klein L; Department of Mathematics and Natural Sciences, University of Applied Sciences, Darmstadt, Germany.; European University of Technology, European Union, Darmstadt, Germany.; Faculty of Medicine, Martin-Luther-University Halle-Wittenberg, Halle (Saale), German., Grieser G; European University of Technology, European Union, Darmstadt, Germany.; Department of Computer Science, University of Applied Sciences, Darmstadt, Germany. |
| Πηγή: | Statistical methods in medical research [Stat Methods Med Res] 2026 Jan; Vol. 35 (1), pp. 79-100. Date of Electronic Publication: 2025 Nov 14. |
| Τύπος έκδοσης: | Journal Article |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: SAGE Publications Country of Publication: England NLM ID: 9212457 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1477-0334 (Electronic) Linking ISSN: 09622802 NLM ISO Abbreviation: Stat Methods Med Res Subsets: MEDLINE |
| Imprint Name(s): | Publication: London : SAGE Publications Original Publication: Sevenoaks, Kent, UK : Edward Arnold, c1992- |
| Ιατρικοί όροι (MeSH): | Machine Learning* , Models, Statistical* , Uncertainty*, Kidney Transplantation/mortality ; Humans ; Computer Simulation ; Likelihood Functions ; Probability |
| Περίληψη: | Clinical prediction models are developed to estimate a patient's risk for a specific outcome, and machine learning is frequently employed to improve prediction accuracy. When the outcome is some event that happens over time, binary classifiers can predict the risk at specific time points if right-censoring is addressed by inverse-probability-of-censoring-weighting . Assessing prediction uncertainty is crucial for interpreting individual risks, but there is limited knowledge on how to consider inverse-probability-of-censoring-weighting when estimating this uncertainty. We propose an adjustment of the infinitesimal jackknife estimator for the standard error of predictions that incorporates inverse-probability-of-censoring-weighting. By using a nonparametric approach, it is broadly applicable, especially to machine learning classifiers. For a simple tractable example, we show that the proposed adjustment reveals unbiased standard error estimates. For other situations, we evaluate performance through simulation studies under both parametric models with inverse-probability-of-censoring-weighting-customized log-likelihood and machine learning with inverse-probability-of-censoring-weighting-customized loss function. We illustrate the methods by predicting post-transplant survival probabilities, using national kidney transplant registry data. Our findings show that the proposed estimator is useful for quantifying prediction uncertainty of inverse-probability-of-censoring-weighting classifiers. Applications to simulated and real data show that prediction uncertainty increases when employing binary classifiers on dichotomized data compared to predictions from survival models. |
| Competing Interests: | Declaration of conflicting interestThe authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Data availability statementAn implementation of the applied analysis methods is available as an R package on GitHub: https://github.com/IDEN-Project-UAS-Darmstadt/IPCWJK. The data analysis presented in this article uses retrospective and anonymized data of the German Transplant Registry. These data have been supplied by the Transplant Registry Agency represented by the Gesundheitsforen Leipzig GmbH. Data can be requested for research purposes here https://transplantations-register.de/en/research. |
| References: | BMC Med Inform Decis Mak. 2021 Feb 6;21(1):43. (PMID: 33549087) J Mach Learn Res. 2014 Jan;15(1):1625-1651. (PMID: 25580094) Kidney Int. 2023 May;103(5):936-948. (PMID: 36572246) Clin Transplant. 2021 Aug;35(8):e14388. (PMID: 34155697) Lifetime Data Anal. 2023 Apr;29(2):441-482. (PMID: 35799026) Stat Med. 2016 May 30;35(12):2016-30. (PMID: 26712471) Transplantation. 2009 Jul 27;88(2):231-6. (PMID: 19623019) Lancet Digit Health. 2021 May;3(5):e295-e305. (PMID: 33858815) Stat Methods Med Res. 2019 Dec;28(12):3755-3768. (PMID: 30514179) Sci Rep. 2022 Nov 14;12(1):19525. (PMID: 36376402) BMC Med Res Methodol. 2018 Feb 26;18(1):24. (PMID: 29482517) NPJ Digit Med. 2022 Jul 11;5(1):89. (PMID: 35817953) Stat Methods Med Res. 2017 Aug;26(4):1654-1670. (PMID: 25934643) Epidemiology. 2010 Jan;21(1):13-5. (PMID: 20010207) Stat Med. 2024 Mar 30;43(7):1384-1396. (PMID: 38297411) Ann Intern Med. 2015 Jan 6;162(1):W1-73. (PMID: 25560730) Stat Med. 2023 Dec 30;42(30):5657-5675. (PMID: 37947168) Transplant Proc. 2020 Jan - Feb;52(1):246-258. (PMID: 31926745) Kidney Int. 2023 Nov;104(5):1035-1036. (PMID: 37863623) J Biomed Inform. 2016 Jun;61:119-31. (PMID: 26992568) Stat Med. 2019 May 20;38(11):2074-2102. (PMID: 30652356) |
| Contributed Indexing: | Keywords: Prediction; inverse-probability-of-censoring; machine learning; survival analysis; uncertainty |
| Entry Date(s): | Date Created: 20251114 Date Completed: 20260109 Latest Revision: 20260112 |
| Update Code: | 20260130 |
| PubMed Central ID: | PMC12783380 |
| DOI: | 10.1177/09622802251393626 |
| PMID: | 41236860 |
| Βάση Δεδομένων: | MEDLINE |
καταχωρήστε σχόλιο πρώτοι!