Don't Let Your Analysis Go to Seed: On the Impact of Random Seed on Machine Learning-based Causal Inference.
| Autor: | Schader L; From the Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University, Atlanta, GA., Song W; From the Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University, Atlanta, GA., Kempker R; Department of Medicine, Division of Infectious Disease, Emory University School of Medicine, Atlanta, GA., Benkeser D; From the Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University, Atlanta, GA. |
|---|---|
| Jazyk: | angličtina |
| Zdroj: | Epidemiology (Cambridge, Mass.) [Epidemiology] 2024 Nov 01; Vol. 35 (6), pp. 764-778. Date of Electronic Publication: 2024 Aug 16. |
| DOI: | 10.1097/EDE.0000000000001782 |
| Abstrakt: | Machine learning techniques for causal effect estimation can enhance the reliability of epidemiologic analyses, reducing their dependence on correct model specifications. However, the stochastic nature of many machine learning algorithms implies that the results derived from such approaches may be influenced by the random seed that is set before model fitting. In this work, we highlight the substantial influence of random seeds on a popular approach for machine learning-based causal effect estimation, namely doubly robust estimators. We illustrate that varying seeds can yield divergent scientific interpretations of doubly robust estimates produced from the same dataset. We propose techniques for stabilizing results across random seeds and, through an extensive simulation study, demonstrate that these techniques effectively neutralize seed-related variability without compromising the statistical efficiency of the estimators. Based on these findings, we offer practical guidelines to minimize the influence of random seeds in real-world applications, and we encourage researchers to explore the variability due to random seeds when implementing any method that involves random steps. Competing Interests: The authors report no conflicts of interest (Copyright © 2024 Wolters Kluwer Health, Inc. All rights reserved.) |
| Databáze: | MEDLINE |
| Externí odkaz: |
|