THE BASE INFORMATION TECHNOLOGY OF HYBRID PATTERNS RECOGNITION
FOR HETEROGENEOUS DATA PROCESSING IN COMPLEX SYSTEMS
- I. Zakhozhai
Volodymyr Dahl East Ukrainian National University
Abstract. The article is dedicated to creation of basic information technology of hybrid recognition for data processing and decision-making in complex systems. The main emphasis is placed on systems that are characterized by the presence of a large number of heterogeneous signs with different nature of origin. At the same time, the data processing is complicated by the fact that the patterns of a complex system have different information content under different current observation conditions also the manifestations of heterogeneous interference and distortion. There are methods and models for processing heterogeneous data in complex systems, but there is no systematic approach to developing of hybrid patterns recognition information technology. The article is devoted to solving this problem. The developed information technology allows to obtain a classification solution with a priori level of reliability. It based on processing a smaller number of the most informative signs. This creates possibility to decline time complexity of data processing. The processing is based on queuing heterogeneous data and then searching for groups of identical classifications. In the case of obtaining a multi-alternative result, classification results are ranked in order their reliability for further expert evaluation. Basic information technology, that was created, may be use for develop of the applied information technologies of hybrid patterns recognition for consolidate data processing in different complex systems.
Keywords: data processing in complex systems, hybrid patterns recognition, patterns recognition reliability increasing, selection of informative signs, reducing the time complexity of the decision-making process.
