چکیده مقاله
Temperature is a crucial parameter to consider due to its impact on daily activities Additionally, it is influenced by other factors such as humidity, dew point, pressure, sunshine duration, and wind speed in the surrounding area Data was obtained from Iran Meteorological Organization IRIMO , from 2015 2021 This study evaluates models for predicting the next day's temperature every hour using Recurrent Neural Networks RNN , Long Short Term Memory LSTM , Gated Recurrent Unit GRU , and 1D Convolutional Neural Networks CNN1D across four scenarios: data from one day, two days, three days, and one week prior Before being predicted, preprocessing is needed to improve data quality consisting of interpolation, sequence formation, encoding time information, and normalization The results of this study prove that using the GRU model produces the best testing MSE, MAPE, and SMAPE values of 3 27, 96 78%, and 27 16% for test data
کلیدواژهها
نویسندگان
Parisa Hormozzadeh
Department of Engineering, Ferdowsi University of Mashhad, Mashhad, Iran
Alireza Shadman
Department of Engineering, Ferdowsi University of Mashhad, Mashhad, Iran
شیوه ارجاع
Hormozzadeh, Parisa and Shadman, Alireza,1403,Hourly Temperature Forecasting Using Deep Neural Networks: A Case Study of Mashhad City,The 10th International Conference on Industrial and Systems Engineering,Mashhad
ارائهشده در
مجموعه مقالات دهمین کنفرانس بین المللی مهندسی صنایع و سیستم ها28 شهریور 1403 · مشهد