Research on Economic Growth Forecasting Models Based on Big Data

Traditional economic growth forecasting methods, though somewhat successful, have limitations such as assuming a linear data generation process and ignoring the nonlinear relationships between economic variables. This study aims to integrate big data technology and machine learning methods to construct an efficient and accurate economic growth forecasting model. By systematically collecting and processing multi-source data, we apply machine learning algorithms such as linear regression, support vector machines (SVM), random forests, and gradient boosting decision trees (GBDT) for modeling and evaluation. The experimental results show that GBDT performs the best across various evaluation metrics, with the lowest mean squared error (MSE) and mean absolute error (MAE), as well as the highest R² value. Further validation through 5-fold cross-validation and time series split methods demonstrates the robustness and generalization ability of the model, showing that the GBDT model outperforms other models in different data partitions and time series predictions.