Abstract:To address the issue of unpredictable rockbursts in engineering, a combined method of Principal Component Analysis (PCA) and Decision Tree Algorithm (DT) is used to predict rockburst levels. The selected predictive indicators include excavation depth H(m), maximum tangential stress (σθ), uniaxial compressive strength (σc), uniaxial tensile stress (σt), stress concentration factor (SCF), brittleness index B1, brittleness index B2, and elastic energy index (Wet). A large sample database is established to create a PCA-DT C4.5 rockburst prediction model. The results indicate that the model achieves an accuracy of 95%, a recall rate of 97.1%, a precision of 95.7%, and an F1 score of 0.964. The model is then validated and applied to the Xin Cheng gold mine project, and its predictive results align with actual conditions. This demonstrates a highly efficient and accurate new method for rockburst prediction.