Systems and Means of Informatics
2019, Volume 29, Issue 4, pp 4-13
DIMENSIONALITY REDUCTION FOR MIXTURE OF PROBABILISTIC PRINCIPAL COMPONENT ANALYZERS IN RELATION TO THE TASKS OF MEDICAL DIAGNOSTICS
Abstract
The article considers algorithms of choosing structural parameters characterizing the mixture of probabilistic principal component analyzers model in relation to the tasks of medical diagnostics. The solution is proposed to search by combining the application of information criteria for the formation of initial approximations, followed by refinement of the resulting solutions. The described approaches and algorithms lead to results that generally do not guarantee the best solution. But they make it possible to clarify whether it is possible to reduce the dimensionality which leads to an increase in the quality of classification. In addition, new information about the objects of study is being formed. Using the example of experiments to diagnose liver diseases and predict the chemical composition of urinary stones, the capabilities of the described data analysis procedures are demonstrated. The proposed solutions are a source of improving the accuracy of classification and give impetus to experts in the subject area to clarify the essence of the processes.
[+] References (6)
- Jensen, R., and O. Shen. 2008. Computational intelligence and feature selection. Rough and fuzzy approaches. Hoboken, NJ: John Wiley & Sons. 340 p.
- Krivenko, M.P. 2019. Vybor razmernostey dlya smesi veroyatnostnykh analizatorov glavnykh komponent [Selecting the dimensionality for mixture of probabilistic principal component analyzers]. Sistemy i Sredstva Informatiki - Systems and Means of Informatics 29 (3):4-15.
- Tipping, M. E., and C. M. Bishop. 1999. Mixtures of probabilistic principal component analyzers. Neural Comput. 11(2):443-482.
- Alber, A. 1987. Multivariate interpretation of clinical laboratory data. New York, NY: CRC Press. 386 p.
- Krivenko, M. P. 2017. Obuchaemaya klassifikatsiya nepolnykh klinicheskikh dannykh [Supervised learning classification of incomplete clinical data]. Informatika i ee Prime- neniya - Inform. Appl. 11 (3):27-33.
- Krivenko, M. P. 2016. Kriterii znachimosti otbora priznakov klassifikatsii [Significance tests of feature selection for classification]. Informatika i ee Primeneniya - Inform. Appl. 10(3):34-40.
[+] About this article
Title
DIMENSIONALITY REDUCTION FOR MIXTURE OF PROBABILISTIC PRINCIPAL COMPONENT ANALYZERS IN RELATION TO THE TASKS OF MEDICAL DIAGNOSTICS
Journal
Systems and Means of Informatics
Volume 29, Issue 4, pp 4-13
Cover Date
2019-11-30
DOI
10.14357/08696527190401
Print ISSN
0869-6527
Publisher
Institute of Informatics Problems, Russian Academy of Sciences
Additional Links
Key words
principal components analysis; Gaussian mixture model; dimensionality reduction; information criterions; cross-validation; medical diagnostics
Authors
M. P. Krivenko
Author Affiliations
Institute of Informatics Problems, Federal Research Center "Computer Science
and Control", Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
|