Publishing house Radiotekhnika

"Publishing house Radiotekhnika":
scientific and technical literature.
Books and journals of publishing houses: IPRZHR, RS-PRESS, SCIENCE-PRESS

Тел.: +7 (495) 625-9241


Analysis of efficiency of the bagging for binary classification at technical diagnostics


Yu.E. Kuvayskova – Ph.D.(Eng.), Associate Professor, Department «Applied Mathematics and Informatics», Ulyanovsk State Technical University
V.N. Klyachkin – Dr.Sc.(Eng.), Professor, Department «Applied Mathematics and Informatics», Ulyanovsk State Technical University

For ensuring reliable and safe functioning of objects, technical diagnostics of their state is carried out. The problem of technical diag-nostics of objects can be interpreted as a problem of binary classification that is reference of technical condition of an object to one of two classes: serviceable or faulty. The process of classification consists in constructing a model that receives the input values of the characteristics of the object, and at the output - the value of the class to which this object belongs. Statistical and mathematical models, models of fuzzy logic and also methods of machine learning can be applied to classification of technical objects. In the article, in order to improve the accuracy of forecasting the technical state of an object, it is proposed to combine the results of binary classification methods using the bagging procedure. In bagging, each classifier is built on the basis of a bootstrap sample. The bootstrap procedure consists in creating a sample that has the same dimensions as the original, and the observations in it are selected randomly from the original sample and can be repeated, while the other elements are absent. In bagging all elementary classifiers learn and work independently of each other, and the decision about the belonging of an object to one of the classes is calculated as the average collective forecast. The efficiency of the proposed technology is demonstrated by the example of a study of a technical object, the state of which is described by eight features. To assess the effectiveness of technical diagnostics of objects, cross-validation with the division of the initial sample into 10 blocks was carried out and the state of the object was classified using the bagging procedure for basic methods of binary classification (discriminant analysis, Bayesian probabilities, logistic regression, decision trees) and without it. To solve the problem, the application package environment MATLAB was used. It is received that the procedure of method bagging allows to increase the accuracy of classification for the considered object by an average of 10.5% in comparison with the basic methods. In this case, the best results are given by the decision tree of decision trees. Just as the averaging of several observations reduces the estimate of the variance of the data, the use of bagging reduces the percentage of the classification error.

  1. Witten I.H., Frank E. Data Mining: Practical Machine Learning Tools and Techniques. San Francisco: Morgan Kaufmann Publishers. 2005. 525 p.
  2. Klyachkin V.N., Kuvajskova Yu.E., Zhukov D.A. Ispol’zovanie agregirovanny’x klassifikatorov pri texnicheskoj diagnostike na baze mashinnogo obucheniya // Sb. trudov III Mezhdunar. konf. i molodezhnoj shkoly’ «Informaczionny’e texnologii i nanotexnologii (ITNT-2017)». Samarskij naczional’ny’j issledovatel’skij universitet imeni akademika S.P. Koroleva. 2017. S. 1770−1773.
  3. Klyachkin V.N., Kuvajskova Yu.E., Alekseeva V.A. Statisticheskie metody’ analiza danny’x. M.: Finansy’ i statistika. 2016. 240 s.
  4. Boks Dzh., Dzhenkins G. Analiz vremenny’x ryadov. Prognoz i upravlenie. M.: Mir. 1974. 242 s.
  5. Vasil’ev K.K., Krasheninnikov V.R. Statisticheskij analiz posledovatel’nostej izobrazhenij. M.: Radiotexnika. 2017. 248 s.
  6. Zadeh L.A. Fuzzy Logic // Computational Complexity: Theory, Techniques and Applications / R.A. Meyers (eds). New York: Springer. 2012. P. 1177−1200.
  7. Kuvayskova Y.E. The Prediction Algorithm of the Technical State of an Object by Means of Fuzzy Logic Inference Models // Procedia Engineering. «3rd International Conference «Information Technology and Nanotechnology» (ITNT 2017)». 2017. S. 767−772.
  8. Kuvajskova Yu.E., Alyoshina A.A. Texnicheskaya diagnostika ob’‘ektov s ispol’zovaniem metodov nechetkoj logiki // Radiotexnika. 2017. № 6. S. 32−34.
  9. Kuvajskova Yu.E., Fedorova K.A., Zhukov D.A. Analiz stabil’nosti raboty’ texnicheskogo ob’‘ekta s primeneniem apparata nechetkoj logiki // Sovremenny’e problemy’ proektirovaniya, proizvodstva i e’kspluataczii radiotexnicheskix sistem. 2016. № 1 (10). S. 167−171.
  10. Klyachkin V.N., Karpunina I.N., Kuvajskova Yu.E., Xoreva A.S. Primenenie metodov mashinnogo obucheniya pri reshenii zadach texnicheskoj diagnostiki // Nauchny’j vestnik UVAU GA(I). 2016. T. 8. S. 158−161.
  11. Breiman L. Bagging Predictors // Machine Learning. 1996. V. 24 (2). P. 123−140.
  12. Shitikov V.K., Masticzkij S.E’. Klassifikacziya, regressiya i drugie algoritmy’ Data Mining s ispol’zovaniem R. 2017. 351 s. (e’lektronnaya kniga, URL =
  13. Voronina V.V., Mixeev A.V., Yarushkina N.G., Svyatov K.V. Teoriya i praktika mashinnogo obucheniya. Ul’yanovsk: UlGTU. 2017. 290 s.
June 24, 2020
May 29, 2020

© Издательство «РАДИОТЕХНИКА», 2004-2017            Тел.: (495) 625-9241                   Designed by [SWAP]Studio