Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review
Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review
}
TY - GEN
T1 - Prediction of the Time Series by the Various Types of Artificial Neural Networks by the Example of Different Time Intervals of the Content of Methane in the Atmosphere
T2 - book chapter
AU - Butorova, Anastasia
AU - Buevich, Alexander
AU - Shichkin, Andrey
AU - Sergeev, Aleksandr
AU - Baglaeva, Elena
AU - Sergeeva, Marina
AU - Subbotina, Irina
AU - Vasilev, Julian
N1 - The equipment of the Common Use Center of Arctic Environmental Research of the Institute of Industrial Ecology of the Ural Branch of RAS was used to measure the concentration of greenhouse gases on Bely Island.
PY - 2023/3/18
Y1 - 2023/3/18
N2 - The paper presents a forecast of changes in the methane content in the air in surface layer of the atmosphere. The forecast was made by the models based on the two most common types of artificial neural networks (ANN): A nonlinear autoregressive neural networks with exogenous inputs (NARX) and Elman neural network (ENN). For training, we used the Levenberg–Marquardt learning algorithm. The data were collected upon monitoring the greenhouse gases on Bely Island, Yamal-Nenets Autonomous Okrug, Russia. For the comparison, the three time intervals with the different patterns of changes in methane content were chosen. To assess the prediction accuracy of the models, we used the mean absolute error, mean square error, and the standardized measure of the model prediction error degree—the index of agreement. The model based on the artificial neural network NARX for all simulated intervals was the most accurate.
AB - The paper presents a forecast of changes in the methane content in the air in surface layer of the atmosphere. The forecast was made by the models based on the two most common types of artificial neural networks (ANN): A nonlinear autoregressive neural networks with exogenous inputs (NARX) and Elman neural network (ENN). For training, we used the Levenberg–Marquardt learning algorithm. The data were collected upon monitoring the greenhouse gases on Bely Island, Yamal-Nenets Autonomous Okrug, Russia. For the comparison, the three time intervals with the different patterns of changes in methane content were chosen. To assess the prediction accuracy of the models, we used the mean absolute error, mean square error, and the standardized measure of the model prediction error degree—the index of agreement. The model based on the artificial neural network NARX for all simulated intervals was the most accurate.
UR - http://www.scopus.com/inward/record.url?partnerID=8YFLogxK&scp=85151057834
U2 - 10.1007/978-3-031-21484-4_34
DO - 10.1007/978-3-031-21484-4_34
M3 - Conference contribution
SN - 978-3-031-21483-7
T3 - Springer Proceedings in Mathematics & Statistics
SP - 383
EP - 388
BT - New Trends in the Applications of Differential Equations in Sciences
A2 - Slavova, А.
PB - Springer Cham
ER -
ID: 37097743