Time Series Analysis and Neural Network Optimization of Tourist Destination Heat Prediction Model

Abstract

Accurate prediction of tourist destination heat is significant for effectively managing tourist flow in the tourism industry. However, forecasting the temperature of a tourist destination is challenging, and various factors, like weather conditions, influence it. In this study, we proposed a novel hybrid strategy using time series analysis and neural network optimization to predict tourist destination heat accurately. The study commences with collecting historical tourist visitation data, weather records, and other information. Then, the raw database was preprocessed to make the data effective and reliable for subsequent analysis. Further, the Autoregressive Integrated Moving Average (ARIMA) technique was employed as a time series analysis technique, capturing and identifying the database's underlying patterns and correlations. Consequently, a Fire Hawk Optimizer-based Dense Recurrent Neural Network (FHO-DRNN) was developed and trained using the ARIMA outcomes to predict the tourist destination heat. The proposed framework was implemented in the MATLAB software, and the results are examined in terms of accuracy, Mean Absolute Error (MAE), and computational time. Furthermore, we compared the existing predictive models to validate the proposed model's effectiveness in forecasting tourist destination heat.