Applying neural networks to assess the efficiency of customs control over fissile and radioactive materials

Currently, the approach to assessing the efficiency of technical means applied in the customs control over fissile and radioactive materials is chosen arbitrarily. As a rule, the assessment is based on statistical indicators determining the technical conditions of using technical means in radiation monitoring at a particular place of operation. This approach is conditioned by the logic of the existence of certain standards and conforms to widespread international practice. However, the presence of potential affecting factors, qualitative characteristics, and additional actions to process statistical data that have not been considered in conventional techniques testifies to the inferiority of assessing the efficiency of the activity of subjects under study, involved in the customs control over fissile and radioactive materials. A way to improve the efficiency of customs control over fissile and radioactive materials is the use of neural networks (i.e., developing a model-methodological complex for processing a large amount of data). A model-methodological complex for assessing the efficiency of technical means of customs control over the FRM should be developed, the combination of properties of which allows considering the permissible impact on the customs system, expanding the possibilities of using managerial actions, reducing material and time expenditures, and improving the credibility of assessing the efficiency of customs control over fissile and radioactive materials in a specific customs body. The paper provides the materials of the developed project of a model-methodological complex for assessing the efficiency of technical means used in the customs control over fissile and radioactive materials.