Open AccessCONSTRUCTION OF A DATA CLUSTERING MODEL EXEMPLIFIED BY DEMO-GRAPHIC INDICATORS OF THE FEFD REGIONS
Author(s) -
Nikita Bezrukov,
AUTHOR_ID,
E. V. Polyanskaya,
AUTHOR_ID
Publication year - 2021
Language(s) - English
DOI - 10.22250/isu.2021.70.3-12
Subject(s) - cluster analysis , computer science , classifier (uml) , basis (linear algebra) , embedding , data mining , artificial intelligence , machine learning , mathematics , geometry
The article deals with the problem of constructing a model for classifying the regions of the Far Eastern Federal District on the basis of demographic data with the use of machine learning algo-rithms - t-distributed Stochastic Neighbor Embedding, K-means and self-organizing networks. Column diagrams and heat maps of correlation coefficients are built for demographic indicators. It is proposed to replace demographic indicators with rank values. The effect it has on the classi-fication results is studied. The classifier has been built on the basis of a self-organizing network, that allows the regions of the Far Eastern Federal District to be classified as belonging to one of the classes: depressed, satisfactory or good.