Detection of Anomalous Components in Spatial Surveys Based on a Multidimensional Model of Poisson Flows and their Cognitive Visualization

Abstract

The paper proposes a technique for detecting anomalous components in spatial scans of multidimensional data in the tasks of multidimensional reviews in GIS technologies. Detection of anomalous components is carried out on the basis of unbiased algorithms under conditions of deep a priori uncertainty regarding the parameters of the distributions of survey data. The results of the detection are controlled by means of cognitive computer graphics. The methods are used to process multidimensional data of astronomical observations. These methods are very successfully applied in astrophysics and can be used for a wide range of tasks in BIG DATA. The methodology of such a combination can also be focused on the identification and forecasting of emergency situations in complex systems.