Open AccessAbout one of the methods of solution of a cluster analysis during illegible source information
Author(s) -
Samira Jalilova
Publication year - 2019
Publication title -
elmi xəbərlər
Language(s) - English
Resource type - Journals
eISSN - 2789-4614
pISSN - 2789-4606
DOI - 10.54414/uesb4172
Subject(s) - closeness , intuition , computer science , taxonomy (biology) , graph , artificial intelligence , cluster (spacecraft) , graph theory , theoretical computer science , machine learning , data mining , pattern recognition (psychology) , mathematics , cognitive science , psychology , combinatorics , mathematical analysis , botany , biology , programming language
Mathematic modeling is still an art, but a quality of models significantly depends on intuition and imagery of their developers. One of the methods of solving mathematic modeling problem is creating a system of recognition based on accumulated information. It is known that two types of problems of the recognition can be distinguished in the mathematic theory: recognition of images with a teacher and recognition without a teacher (problems of cluster analysis, taxonomy, automatic classification or simply classification). This work focuses on the classification problem with respect to the subgraphs of this graph which determines the status of complex systems. A theorem on metrics, which characterizes the “closeness” of the status of complex systems, has been proved.