Applications of Deep Learning to Predict Epidemics in the Kingdom of Saudi Arabia Based on Time Series Data: A Comparative Study

Abstract

The current study aimed at identifying the applications of deep learning to predict epidemics in the Kingdom of Saudi Arabia based on time series data. Time series analysis paid great attention to prediction as many studies indicated. Also, many applications were proposed for the purpose of predicting the future of life phenomena, diseases and epidemics. Due to the importance of comparison among the different prediction methods, the aim of this study was to make a comparison between the long-short-term memory network model and the Box-Jenkins Model in predicting infection or death resulting from epidemics in the Kingdom of Saudi Arabia. Each model was tested to predict three new days and a comparison was made between them based on relevant statistical criteria. The relationship such as (RMSE), (RUNS), (VAR), (MEAN) (AUTO) (RUNM). The study concluded that the proposed model that is appropriate for time series data is the model of neural networks with long short-term memory, and it has been applied to time series data. For the numbers infected with epidemics in the Kingdom of Saudi Arabia. Each model was tested to predict three new days and a comparison was made between them based on relevant statistical criteria such as (RMSE), (RUNS), (VAR), (MEAN), (AUTO) (RUNM). The study concluded that the proposed model that is appropriate for time series data is a model of neural networks with long-short-term memory. It was applied to time series data of numbers infected with epidemics in the Kingdom of Saudi Arabia.