The problems of solving multi-criteria modeling tasks in conditions of heterogeneous data uncertainty are characteristic and relevant for many areas, in particular, transport and logistics, medical, and economic. This paper summarizes the approaches to solving these ar-eas tasks, including cases for data obtained during monitoring with uneven and unclear inter-vals. All the considered approaches were created under the guidance and with the direct par-ticipation of prof. Skalozub V. V. The author's team, with his participation, illustrated exam-ples of the use of models in the tasks of organizing multi-symbol sequences in transport (for-mation-dismantling of trains, order processing) and in the analysis and forecasting of states/events based on data on diabetes diseases. In this paper, the main approaches to modeling in conditions of heterogeneous data un-certainty based on multi-layer constructive-synthesizing, separable and relational-separable models are highlighted. Multi-layer constructive-synthesizing became new branch of develop constructive-synthesizing modeling. Their practical value currently lies in the possibility of obtaining a new form of implementation of the specified technological processes of railway transport. In the future, they can be applied to tasks that require step-by-step data processing with a clear demarcation of the relevant operations. The obtained results can be used for further development of the proposed ideas and ap-proaches. It is promising to develop the identified approaches, including by combining them, as well as supplementing constructive-production models with a formalized description of con-structive elements, including input data, by ontological means. The paper contains information about the figure and scientific work of V. Skalozub.
Read full abstract