| Autor: |
Yeremeyev, I., Dychko, A., Kyselov, V., Remez, N., Kraychuk, S., Ostapchuk, N. |
| Předmět: |
|
| Zdroj: |
Latvian Journal of Physics & Technical Sciences; Aug2021, Vol. 58 Issue 4, p69-78, 10p |
| Abstrakt: |
The present paper provides the ways of implementing methods of fuzzy set approach, which contributes to an increase in the accuracy, efficiency and functional flexibility of the complex control and recognition monitoring systems for the environmental and geoengineering system simulation. They are based on data mining methods and may be implemented with the help of intellectual technologies, including the combination of model pluralism, membership functions, methods of nearest neighbour, results of fractal and chaos theories, methods of ensuring robustness and retrospective analysis of the decision tree for success in decision making in similar situations. It is proposed to use the model pluralism to explain a particular information process, which uses a number of adequate models, describing the behavior of objects in the case where each of the model reflects its behavior objectively, but under different circumstances, which are difficult to consider a priori in real time when choosing an adequate model. It is shown that the method of the nearest neighbour should be used if it necessary to identify causal relationships and predict further development of the environmental safety events. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
