Assessment of the quality of a binary classifier in research
https://doi.org/10.51523/2708-6011.2020-17-4-15
Abstract
The article deals with the basic principles of the assessment of the quality of a binary classifier using the potential of the ROC analysis and describes a ROC analysis algorithm and the interpretation of the obtained results. The work discusses the methodology for the calculation of the main parameters of the ROC curve, as well as the instructions for the assessment of their statistical significance.
About the Authors
A. A. Kovalev
Gomel State Medical University
Belarus
Belarus
Aleksey A. Kovalev — Head of the Sector of Operational Maintenance and Repair of Equipment of the EI «Gomel State Medical University»
B. K. Kuznetsov
Gomel State Medical University
Belarus
Belarus
Boris K. Kuznetsov — Candidate of Biological Science, Associate Professor, Head of the Department of Medical and Biological Physics of the EI «Gomel State Medical University»
A. A. Yadchenko
Gomel State Medical University
Belarus
Belarus
Aleksey A. Yadchenko — Doctor of Physics and Mathematics, Professor at the Department of Medical and Biological Physics of the EI «Gomel State Medical University»
V. A. Ignatenko
Institute of Biochemistry of Biologically Active Compounds of the National Academy of Sciences of Belarus
Belarus
Belarus
Valeriy A. Ignatenko — Candidate of Biological Science, Head of the Branch Laboratory for Monitoring the Nutritional (Micronutrient) Status of the Population and Development of Technologies for Its Correction Using Functional Products and Biologically Active Supplements, SE «Institute of Biochemistry of Biologically Active Compounds of the National Academy of Sciences of Belarus»
References
1. Fletcher R, Fletcher S, Vagner E. Klinicheskaya epidemiologiya. Osnovy dokazatel'noj mediciny. Moskva, RF: Media Sfera; 1988. 352 p. (in Russ.)
2. Fajnzil'berg LS, ZHuk TN. Garantirovannaya ocenka effektivnosti diagnosticheskih testov na osnove usilennogo ROC-analiza. Upravlyayushchie Sistemy i Mashiny. 2009;5:3-13. (in Russ.)
3. Gajdyshev IP. Ocenka kachestva binarnyh klassifikatorov. Vestnik Omskogo Universiteta. 2016; 1(79): 14-17. https://cyberleninka.ru/article/n/otsenkakachestva-binarnyh-klassifikatorov/viewer (in Russ.)
4. Litvin AA, Kalinin AL, Trizna NM. Ispol'zovanie dannyh dokazatel'noj mediciny v klinicheskoj praktike. Problemy Zdorov'ya i Ekologii. 2008 Vol 18(4):12-19. (in Russ.)
5. Pacaj DI, Blahov NYU. ROC-analiz ocenki tyazhesti sostoyaniya i kachestva prognoza u bol'nyh ostrym pankreatitom. Voennaya Medicina. 2011;42(3):60-66. (in Russ.)
6. David J. Hand. Measuring classifier performance: a coherent alternative to the area under the ROC curve. Machine Learning. 2009;77:103-23. doi: 10.1007/s10994-009-5119-5
7. Seong Ho Park, Jin Mo Goo, Chan-Hee J. Receiver Operating Characteristic (ROC) Curve: Practical Review for Radiologists Korean Journal of Radiology. 2004;5(1):11-18. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2698108/
8. Hanczar B, Hua J, Chao Sima, Weinstein J, Bittner M, Edward R. Dougherty Small-sample precision of ROCrelated estimates. Bioinformatics. 2010; 26(6):822-30. https://doi.org/10.1093/bioinformatics/btq037
9. Korolyuk IP. ROC-analiz (operacionnye harakteristiki nablyudatelya): bazovye principy i primenenie v luchevoj diagnostike. Medicinskaya Vizualizaciya. 2013;6:113-23. (in Russ.)
Review
For citations:
Kovalev A.A., Kuznetsov B.K., Yadchenko A.A., Ignatenko V.A. Assessment of the quality of a binary classifier in research. Health and Ecology Issues. 2020;(4):105–113. (In Russ.) https://doi.org/10.51523/2708-6011.2020-17-4-15
Views: 418
Email this article 