Ranked classification of environmental conditions

Abstract

This paper discusses the main features associated with the development and research of the device based on the methods of intelligent technology for assessing the state of the natural environment. It should be noted that natural and technical objects, as well as the processes occurring in them, are characterized by high complexity and dynamism, and a significant part of these processes has not yet been fully studied and formalized. Therefore, one of the most important areas of data analysis in this area is the use of artificial neural networks in information and measurement systems. In the works of scientists from various countries, the high efficiency of the use of artificial neural networks in solving individual data processing problems in the classification of environmental conditions is shown. The proposed approach is based on methods of nonparametric statistics using rank criteria and will allow for intelligent analysis of data on key environmental indicators, such as hydrometeorological data on the level of pollution and composition of air, soil, maximum permissible emissions of harmful substances, environmental monitoring of anomalies, and others. Static, dynamic, integral, and generalized models of classification of environmental conditions are presented. Further research plans suggest evaluating the impact of sample size on statistical sensitivity, statistical stability, and areas of confident/uncertain recognition, as well as building a decision support system for detecting the G-effect, and considering an adaptive approach to constructing an evaluation matrix.