Petryshyn S. Information technology of clustering states of computer equipment

Українська версія

Thesis for the degree of Candidate of Sciences (CSc)

State registration number

0416U001546

Applicant for

Specialization

  • 05.13.06 - Інформаційні технології

03-03-2016

Specialized Academic Board

Д 05.052.01

Vinnytsia national technical university

Essay

The object of study - the process of clustering states of computer equipment. The aim of the research is to improve the quality of clustering states of computer equipment through the development and implementation of information technology for further decisions on the operation of such equipment. The methods used system analysis and synthesis, intelligent analysis, object modeling process of clustering states of computer equipment, computer simulation. The theoretical results are: first proposed information model of clustering states of computer equipment, which feature is that it is based on the model of states of computer equipment; it takes into account the characteristics of the weighting coefficients and parameters of these states in the process of determining the distance between them; it uses valuation of these parameters and characteristics, thus improving the quality of the clustering; improved clustering method K-MEANS, which, unlike existing account tolerances assumes values of the parameters and characteristics of the computer equipment in the formation of clusters of its states, thus improving the quality of clustering conditions in a known number of clusters; improved clustering method FOREL, which, unlike the existing ones, provides for the calculation of the radius of clusters of states of the computer equipment to reduce emissions from them, thus improving the quality of the partition of the set of states in the absence of information about the number of clusters; first developed an information technology of clustering states of computer equipment, which feature is the proposed use of the information model of the clustering process, as well as improved clustering methods K-MEANS and FOREL, thus improving the quality of partitions of the set of such states. Practical results: improved algorithm K-MEANS of clustering of states of computer equipment was developed; improved algorithm FOREL of clustering of states of computer equipment was developed; developed software of clustering of states of computer equipment. The degree of implementation: research results are implemented in the company LLC "IKrok", in the Department of Energy, transport and communication Vinnytsia city council, as well as in the educational process of Vinnytsia National Technical University in the Department of Computer Science. Scope (industry) of application - maintenance of computer equipment.

Files

Similar theses