INFORMATION ANALYSIS METHOD ABOUT CURRENT SITUATIONS IN ACS OF SPECIAL OPERATIONS

• Main Authors: Maksim Pavlenko, Serhii Khmelevskiy, Oleksii Petrov
• Format: Article
• Language: English
• Published: National Technical University "Kharkiv Polytechnic Institute" 2020-03-01
• Series: Сучасні інформаційні системи
• Subjects: ERP system; recognition methods of situations; information subsystems
• Online Access: http://ais.khpi.edu.ua/article/view/198544

The relevance of improvement and difficulty of establishing ERP systems and methods of their development is considered. The features of the development of special mathematical and software implements strategy ERP system shows effective ways formalized description of the classes of recognizable situations in the form of membership functions corresponding regions of the feature space. In the course of developing automated ERP systems, it is necessary to apply direct methods for characterization of classes of recognizable situations. As shown by the analysis of typical SR tasks solved in ERP systems, in practice a feature space can contain many hundreds and even thousands of attributes. In this case, the number of examples of situations belonging to a particular class, as a rule, is limited to tens of units. Therefore, for classical learning algorithms, the required length of a training sequence for SR tasks of practical importance, as a rule, many times exceeds the disposable number of examples of environment. The source of information necessary for the construction of such a priori descriptions can be experts who can solve SR tasks in a non-automated way in conditions of psychological comfort.