Intellectual, neuronetic and the geoinformation technologies, distributed information systems.

Chief researchers:

Departments:

Basic findings:

1. Hybrid models of knowledge that allow to carry out geomodeling with elements of activization of semantic GIS information are developed. Methods of construction of static and dynamic parallel-series-alternative structures of knowledge which allow to analyze logic connections in productions system are offered. Dynamic structures of knowledge are applied for modeling the development of complex scripts and the analysis of consequences of states of emergency. Object-oriented realization of language of knowledge representation is offered. Algorithmic facilities and software are developed. The technology of construction of expert geoinformation systems is created. Approbation for problems of warning and liquidation of states of emergency is executed. A new approach to the support of organizational management on the basis of OLAP (On-Line Analytical Processing) methods is offered: the decision of administrative problems (analysis, estimation, planning) by means of designing analytical models. Approbation of the offered approach is carried out for the problems of the analysis and planning of volumes of medical aid to the population of Krasnoyarsk Region.[more...].
2. New algorithms of nonparametric processing of information are created[more...]
3. The analysis of the regional trends of climatic parameters in Siberia and Russia is executed: temperatures, deposits, continentality etc. By means of geoinformation system the division of Siberian and Russian territory into districts by the results of research of long-term fluctuations of climatic variables is made. The areas with the greatest changes of climatic variables are allocated. [more...].
4. Fast methods of training neural networks and the technology of extraction of knowledge from training thinning neural networks are created, the software is developed and numerous applied problems in various areas are investigated. [more...].
5. Methods and Neuro-GIS software for visualization and mapping of the data of arbitrary nature are developed. On the basis of Neuro-GIS methods for data visualization and mapping the method of automatic identification of protein — coding sites in genome is developed. [more...].