Neural Network Modeling and Assessment of the Territorial Situation

The technologies of using the apparatus of an artificial neural network in a new subject area - procedures for spatial assessment of the rapidly changing situation in the region are considered. The tasks of neural network modeling of a complex dynamic territorial structure - a regional situation are...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE NW Russia Young Researchers in Electrical and Electronic Engineering Conference s. 514 - 517
Hlavní autoři: Zinchenko, Elena G., Bidenko, Sergey I., Khramov, Igor S., Minakov, Vladimir F., Dytlov, Sergey A., Ivanovskii, Aleksei N.
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 25.01.2022
Témata:
Analytical models
artificial neural network
Artificial neural networks
Backpropagation algorithms
Economics
Learning systems
Multilayer perceptrons
neural network model
Recurrent neural networks
spatial analysis
territorial economic situation
topological image of the geo-image of the economic situation
ISSN:2376-6565
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The technologies of using the apparatus of an artificial neural network in a new subject area - procedures for spatial assessment of the rapidly changing situation in the region are considered. The tasks of neural network modeling of a complex dynamic territorial structure - a regional situation are formulated. The technologies for the formalized representation of the territorial situation in the region for spatial analysis using the apparatus of artificial neural networks have been determined. The order of formation of parameters and the range of output estimates are established. As a basic structure of a neural network, a recurrent neural network with a multilayer perceptron architecture is proposed. A neural network learning mechanism is substantiated in the form of an error backpropagation algorithm, which is optimal for the classification problem using a recurrent neural network.
ISSN:2376-6565
DOI:10.1109/ElConRus54750.2022.9755500