Open AccessForecast of the near ground air temperature based on the multilayer perceptron model
Author(s) -
Irina V. Del,
Alexander V. Starchenko
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1989/1/012025
Subject(s) - multilayer perceptron , artificial neural network , perceptron , mean squared error , moment (physics) , computer science , air temperature , set (abstract data type) , weather forecasting , meteorology , environmental science , data mining , artificial intelligence , statistics , mathematics , geography , physics , classical mechanics , programming language
In this study, a multilayer perceptron model is implemented for predicting meteorological values. Based on the known distribution of meteorological values for several previous days, the task was set to predict the values of the near ground air temperature. The overall mean square error for the entire forecast was 3.11 C. Comparison of various optimization methods showed the advantage of the method of Adaptive Moment Estimation. Comparison of the multilayer perceptron model forecasting results with the Weather Research and Forecasting numerical model forecast showed the promise of using neural networks to predict meteorological parameters at weather observation points.