Hybrid SARIMA-LSTM Model for Tourism Industry in Malaysia

Tourism is one of the fastest growing economic industry in the world, and at the same time become the main income generator for a significant number of developing countries over the decades. Unfortunately, due to the COVID-19 pandemic in March 2020, this industry was severely impacted and almost collapsed. In order to cushion the impact of this outbreak, minimize job crisis and support recovery in 2021 and beyond, an accurate tourism demand forecasting played a vital role nowadays. Therefore, this study focused on tourism demand forecasting using hybrid SARIMA and long-short term memory (LSTM) model. The data used in this study was monthly historical data on the number of tourist arrivals in Malaysia before the onset of the Movement Control Order (MCO) starting from January 2000 to February 2020 due to the COVID-19 outbreak. The main objective of this study was to enhance the tourism demand forecasting accuracy in Malaysia by implementing a hybrid SARIMA model with LSTM model. SARIMA model was implemented to cater the linear part meanwhile LSTM networks, the extension of the recurrent neural networks (RNN) were used to analyze the nonlinear part of the data. The forecast evaluation implemented to predict the error of this hybrid model are Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) and all analyses was performed by using Python software. The results showed that the proposed hybrid SARIMA-LSTM model was considered as highly accurate forecasting model due to the percentage error produced.