UDC 004.724
APPLICATION OF MACHINE LEARNING METHODS TO CLASSIFY MATERIALS BASED ON THEIR KEY CHARACTERISTICS
V. P. Koryachko, Doctor in Technical Sciences, Рrofessor, CAD department, Head of the Department, RSREU,Ryazan, Russia;
orcid.org/0000-0003-0272-673X, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
S. D. Vikulin, post-graduate student, RSREU, Ryazan, Russia;
orcid.org/0009-0002-9932-1113, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
The problem of multiclass classification of materials by their key properties using machine learning methods and neural network models is considered. The aim of this work is to build an effective model for analyzing the properties of materials and for determining their class of affiliation. Special attention is paid to the problem of class imbalance that occurs when classifying rare types of materials. The method of syn thetic increase of minority classes (SMOTE) is used to eliminate imbalance. The quality of the model is as sessed using standard classification metrics, including accuracy, completeness, and F1-measure. The results obtained demonstrate the effectiveness of the proposed approach in the task of classifying materials and the possibility of its application in intelligent decision support systems in materials science.
Key words: : machine learning, classification, neural network models, class imbalance, SMOTE, material properties, intelligent systems.