Processing of Nonlinear Heterogeneous Data in Information and Control Systems

The article presents a method for cluster analysis of heterogeneous data using the provisions of fuzzy logic. Heterogeneous are data of various formats and collected from various sources. In addition, such data is usually incomplete and inaccurate, which makes it difficult to process and cluster them. The article presents mathematical models for representing a candidate for a vacant position, which is characterized by heterogeneous data. The apparatus of algebraic systems was used to develop mathematical models. A method for determining the membership function of fuzzy sets using a probabilistic approach, as the most effective when working with heterogeneous data, is described in detail. An example of the formation of a base of logical rules for the formation and selection of classification features in a set of heterogeneous data of the personnel reserve of a manufacturing enterprise is given. The selected classification features allow for further accurate and efficient verification and evaluation of information about candidates for a vacant position. In addition, the proposed method of cluster analysis of heterogeneous data can be applied in various subject areas that involve the use of incomplete and inaccurate data, for example, socio-economic, technical, biological systems.