Voronezh, Russian Federation
Logistic regression is one of the most widely used methods for analyzing binary outcomes in medicine and biology. The method provides a probabilistic interpretation of the result and allows the contribution of risk factors to be assessed through the odds ratio. The paper proposes a reproducible methodology for constructing and validating a logistic model for medical data, including data preprocessing, model training, and quality assessment using ROC analysis and probability calibration. The practical application is demonstrated on the problem of diabetes mellitus prediction. The obtained results confirm the suitability of logistic regression as an interpretable tool for supporting clinical decision-making.
logistic regression, biostatistics, ROC analysis, AUC, medical data, risk factors
1. Lapach S. N. Statisticheskie metody v mediko-biologicheskih issledovaniyah s ispol'zovaniem Excel Eksperim. issled. Klin. ispytaniya. Analiz farmac. rynka / S.N. Lapach, A.V. Chubenko, P.N. Babich. – 2-e izd., dop., s magnit. nositelem. – Kiev : MORION, 2001. – 407 s. il.; 26. – (Medicina i statistika); ISBN 966-7632-33-4.
2. Saharnyy diabet : diagnostika, lechenie, profilaktika : rukovodstvo / I. I. Dedov [i dr.] ; pod red. I. I. Dedova, M. V. Shestakovoy. – Moskva : Med. inform. agentstvo, 2011. – 801 s. : cv. il., portr. – ISBN 978-5-9986-0061-6.
3. Clinical prediction models: A practical approach to development, validation, and updating Ewout W. Steyerberg (2019). Second Edition, Springer Series Statistics for Biology and Health. Cham: Springer. 558 pages. ISBN: 978-3-030-16398-3. https://doi.org/10.1007/978-3-030-16399-0.
4. Kim J. H. Multicollinearity and misleading statistical results // Korean Journal of Anesthesiology. – 2019. – Vol. 72, No. 6. – P. 558–569. DOI: https://doi.org/10.4097/kja.19087
5. Lachiche N., Flach P. Improving accuracy and cost of two-class and multi-class probabilistic classifiers using ROC curves // Proceedings of the Twentieth International Conference on Machine Learning (ICML-2003). – Washington : AAAI Press, 2003. – Vol. 20. – P. 416–424.
6. Kuhn M., Johnson K. Applied Predictive Modeling. – New York : Springer, 2013. – 600 p. DOI: https://doi.org/10.1007/978-1-4614-6849-3
7. Korobova L. A., Eynullaev T. A. O., Ten R. V. Sistemnyy podhod pri reshenii prikladnyh medicinskih diagnosticheskih zadach // Modelirovanie sistem i processov. – 2024. – T. 17, № 3. – S. 52–61. – DOI:https://doi.org/10.12737/2219-0767-2024-50-59.
8. Arapov D. V., Kuricyn V. A., Skorobogatov S. A. Avtomatizirovannaya mobil'naya sistema prinyatiya resheniy dlya ispol'zovaniya v special'noy medicine // Modelirovanie sistem i processov. – 2024. – T. 17, № 3. – S. 7–16. – DOI:https://doi.org/10.12737/2219-0767-2024-5-13.
9. Yurchishina M. V., Bushmeleva K. I. Algoritmicheskaya model' SPPR «Optimal'nyy uchebnyy plan» // Modelirovanie sistem i processov. – 2024. – T. 17, № 4. – S. 84–95. – DOI:https://doi.org/10.12737/2219-0767-2024-17-4-84-95.
