Research on Wildlife Trade Forecasting Based on Linear Regression and ARIMA Models

This study developed a forecasting model for predicting the trends in global illegal wildlife trade. The model is based on historical data and integrates linear regression and ARIMA models, with a weighted optimization method to enhance forecasting accuracy. In the data preprocessing stage, Newton's interpolation method was used to fill in missing values, ensuring data integrity. Furthermore, a comprehensive index system was constructed, and the impact of these factors on illegal trade was assessed using a multiple linear regression model. Additionally, linear regression analysis was conducted using MATLAB, and after confirming the stationarity of the time series data through the ADF unit root test and Eviews software, an ARIMA model was established. Ultimately, an optimization algorithm was used to determine the weights of each model, achieving a weighted prediction. The model evaluation employed RMSE, MAPE, and R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> as indicators, and the results showed that the model has high predictive accuracy and stability.