THE METHOD OF MULTILAYER CLASSIFICATION FOR PREDICTING POSTOPERATIVE COMPLICATIONS
https://doi.org/10.18705/2311-4495-2014-0-4-11-18
Abstract
This article proposes a method of improving the classification of postoperative complications (postpericardiotomic syndrome and low ejection fraction) on the basis of regularized discriminant analysis applied to several sub-groups with more favorable classification properties. Logical operations and finite linear combinations over F2 in set of factors are used for identification of recent. Prediction is obtained by averaging of co-incident discriminant functions and by frequency of their correct classification.
About the Authors
Nina P. Alekseeva
Federal Almazov Medical Research Centre; Saint Petersburg State University; Pavlov First Saint Petersburg State Medical University
Russian Federation
Russian Federation
A. O. Andreev
Federal Almazov Medical Research Centre
Russian Federation
Russian Federation
P. V. Ananyevskaya
Federal Almazov Medical Research Centre
Russian Federation
Russian Federation
L. A. Belyakova
Federal Almazov Medical Research Centre; Pavlov First Saint Petersburg State Medical University
Russian Federation
Russian Federation
D. M. Komleva
Saint Petersburg State University
Russian Federation
Russian Federation
M. M. Kocheryzhkina
Federal Almazov Medical Research Centre
Russian Federation
Russian Federation
O. M. Moiseeva
Federal Almazov Medical Research Centre
Russian Federation
Russian Federation
E. V. Nakatseva
Federal Almazov Medical Research Centre
Russian Federation
Russian Federation
D. V. Tchuev
Federal Almazov Medical Research Centre
Russian Federation
Russian Federation
B. B. Bondarenko
Federal Almazov Medical Research Centre
Russian Federation
Russian Federation
References
1. Cleophas T.J., Zwinderman A.H. Machine Learning in Medicine. - 2013. - XV. - 265 p.
2. Kun Zh., Ying-Jie T., Nai-Yang D. Robust Unsupervised and Semi-supervised Bounded nu -Support Vector Machines // Advances in Neural Networks. Lecture Notes in Computer Science. - 2009. - Vol. 5552. - P. 312-321.
3. Schapire R.E., Freund Y. Boosting: Foundations and Algorithms. Adaptive Computation and Machine Learning series. - MIT Press, 2012.
4. Friedman J.H. Stochastic gradient boosting // Comput. Stat. Data Anal. - 2002. - Vol. 38, № 4. - P. 367-378.
5. Friedman J., Hastie T., Tibshirani R. Additive logistic regression: a statistical view of boosting // Ann. of Stat. - 2000. - Vol. 28, № 2. - P. 337-407.
6. Roe B.P., Yang H.J., Zhu J. et al. Boosted decision trees as an alternative to artificial neural networks for particle identification // Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment. - 2005. - Vol. 543, Issue 2-3. - P. 577-584.
7. Zhu J., Hastie T. Classification of gene microarrays by penalized logistic regression // Biostat. - 2004. - Vol. 5, № 3. - P. 427-443.
8. Asafu-Adjei J.K., Sampson A.R., Sweet R.A., Lewis D.A. Adjusting for matching and covariates in linear discriminant analysis // Biostatistics. - 2013. - Vol. 14, № 4. - P. 779-791.
9. Guo Y., Hastie T., Tibshirani R. Regularized linear discriminant analysis and its application in microarrays // Biostatistics. - 2007. - Vol. 8, № 1. - P. 86100. - Epub: 2006, Apr 7.
10. Алексеева Н.П. Анализ медико-биологических систем. Реципрокность, эргодичность, синонимия. - Санкт-Петербург: Издательство Санкт-Петербургского университета, 2012. - 184 с.
11. Алексеева Н.П., Иванова Е.П., Бондаренко Б.Б. и др. О способе улучшения прогнозирования вазоспазма лучевой артерии на основе симптомного расслоения популяций. // Ученые записки СПбГМУ им. акад. И.П. Павлова, - 2009. - T. XV, № 4. - С. 59-62.
12. Елькин А.В., Титаренко О.Т., Алексеева Н.П. и др. Оценка риска послеоперационных осложнений у больных фиброзно-кавернозным туберкулезом легких // Проблемы туберкулеза и болезней легких. - 2009. - № 5. - С. 31-34.
13. Титаренко О.Т., Эсмедляева Д. С., Алексеева Н.П. и др. Сравнительная ценность биохимических маркеров клеточного иммунитета в диагностике туберкулезного плеврита // Клиническая лабораторная диагностика. - 2010. - № 1. - С. 46-49.
14. Алексеева Н.П., Алексеев А.О., Бондаренко Б.Б. Симптомно-синдромальный подход в исследовании информативности комплекса категориальных признаков // Артериальная гипертензия. - 2008. - T. 14, № 1. - С. 25.
15. Алексеева Н.П., Конради А.О., Бондаренко Б.Б. Симптомный анализ в исследовании долгосрочного клинического прогноза // Артериальная гипертензия. - 2008. - T. 14, № 1. - С. 38-43.
Review
For citations:
Alekseeva N.P., Andreev A.O., Ananyevskaya P.V., Belyakova L.A., Komleva D.M., Kocheryzhkina M.M., Moiseeva O.M., Nakatseva E.V., Tchuev D.V., Bondarenko B.B. THE METHOD OF MULTILAYER CLASSIFICATION FOR PREDICTING POSTOPERATIVE COMPLICATIONS. Translational Medicine. 2014;(4):11-18. (In Russ.) https://doi.org/10.18705/2311-4495-2014-0-4-11-18
Views:
704
Email this article 