VARIMA and the RNN weather predictions Comparison in Irbid city
Abstract
This research aims to compare two weather forecasting models: the Au- toregressive Integrated Moving Average Model and the Recurrent Neural Network model. Analysis and forecasting of boundary layer variables in the city of Irbid is performed based on historical data for many weather variables. The performance of both models is evaluated using the root mean square pre- diction error (normalized RMSPE) to predict each variable separately. The results showed that the Recurrent Neural Network model was superior to the Autoregressive Integrated Moving Average Model in predicting five variables out of the six variables used in the study.
DOI Code:
10.1285/i20705948v17n2p317
Keywords:
Weather prediction, LSTM, Time Series, deep learning, Machine learning
Full Text: pdf