APPLICATION OF MACHINE LEARNING CLASSIFIERS IN DETERMINING SIGNAL QUALITY IN PARAMETRIC TECHNOLOGICAL PROCESSES

Ilhomjon Qodirov

Authors

       

   Ibragimov U.M. – PhD, Associate Professor, Bukhara Institute of Engineering and Technology, Bukhara, Uzbekistan, ciulugbek@list.ru

Annotation

        This article discusses the basic concepts and methods used in solving the classification problem in machine learning. The definition of classification as the process of assigning objects to predefined classes based on the analysis of object features is given. Exam-ples of binary and multi-class classification are given. On the example of constructing the sim-plest signal classifier using a decision tree, the process of creating a classifier and analyzing its effectiveness based on test data using a confusion matrix and quality indicators such as accura-cy, completeness, F-measure is examined in detail. Formulas for calculating these indicators are given. The importance of representativeness of test data for assessing the quality of the classifier is noted. In conclusion, the author concludes that the accuracy on the test set Te is determined as access only to the true classes of a small part of the instance space, therefore, the estimate is everything we can hope for. Therefore, it is important that the test set be as repre-sentative as possible. This is usually formalized by the assumption that the appearance of in-stances in the world is governed by an unknown probability distribution on X, and that the test set Te is generated according to this distribution. In general, the article is of interest to special-ists in the field of machine learning and data analysis.

Key words

  classification, machine learning, function, signal filtering, binary, feature tree, test instances, probability

Language

english

Type

technical

Year

2023

Page

24-25

References

 

    1. Avezov A.Kh. Application of information technologies for teaching scientific creativity. Bulletin of the Tajik State University of Law, Business and Politics. 2014. No. 1 (57). pp. 316-327.
    2. Allwein, E.L., Schapire, R.E. and Singer, Y. (2000). Reducing multiclass tobinary: A unifying approach for margin classifiers. In P. Langley (ed.), Proceedings of the Seventeenth International Conference on Machine Learning (ICML 2000), pp. 9–16.
    1. Blockeel, H., De Raedt, L. and Ramon, J. (1998). Top-down induction of clustering trees. In J.W. Shavlik (ed.), Proceedings of the Fifteenth International Conference on Machine Learning (ICML 1998), pp. 55–63.
    1. Boser, B. E., Guyon, I. and Vapnik, V. (1992). A training algorithm for optimal margin classifiers. In Proceedings of the International Conference on Computational Learning Theory (COLT 1992), pp. 144–152.
    1. Breiman, L., Friedman, J.H., Olshen, R.A. and Stone, C.J. (1984). Classification and Regression Trees. Wadsworth.

Publication date

2023-10-19