Cartography and geoinformatics
UDC: 
Year: 
2016
Tom: 
21
#: 
4
36
Pages: 
74-88
Abstract:
The possibility of constructing a system of space monitoring using data mining techniques such as association and classification methods based on Bayesian decision rule using methods on the basis of the adoption of decision trees are discussed. Examples of regional aerospace monitoring problems (environmental monitoring, regional forest fire monitoring, monitoring of local natural and anthropogenically provoked by events and disasters, and others.) and the corresponding data analysis techniques (regression, anomaly detection, spatial prediction, etc.) with the support of decision-making are discussed. The problems of adaptation of these methods to the archives of remote sensing data, involving as accounting features of data analysis methods, and specific aspects of building monitoring systems are presented. We discuss the example of a structure of transactional data for aerospace monitoring system comprising a set of reading meters (sensors) of a physical quantity, which is valid for sensor placement location in a spatial neighborhood. The approaches to the use of data analysis in systems monitoring using the optional archive transactional data such structure on the state of the environment, formulated conceptual requirements for such a system, its structure and the generalized use of technology to unmanned aerial vehicles.
Free access. Download PDF: 
Publication History
Received: 23.09.2016
Published: 19.12.2016

Content

2016 issue 4