Comparative Analysis of Time Series Models and Panel Data Models in Economic Growth Forecasting

This paper mainly shows the comparison of two models, the time series model and the panel data model in economic data processing and forecasting. Using GDP data from the United States, Japan, and the United Kingdom, the study assesses the advantages and limitations of each model in predicting economic trends. In terms of time series models, the ARIMA model (AutoRegressive Integrated Moving Average) is utilized for single-variable, non-stationary data analysis, offering robust short-term forecasts while accounting for trends and seasonality. But the disadvantage is that it cannot handle more complex data and cannot show the impact of other factors on the data being processed. On the other hand, the panel data model efficiently handles multi-dimensional datasets, incorporating economic indicators from various countries to explore inter-variable relationships. The disadvantages are that it requires a huge database, greater time cost, etc. Although the fit to history is good, the prediction results are not very accurate without a good simulation environment. Through this comparison, comparison of relevant statistical values provides a certain reference for selecting models for future economic predictions. Evaluation metrics such as Mean Squared Error (MSE), Mean Absolute Error (MAE), and R-squared (R) were used to measure the models accuracy.