Machine Learning Method Predicts Risk Level of Coal Mine Roof Accidents

Researchers at the China University of Mining and Technology have developed a novel method for predicting the risk level of coal mine roof accidents using machine learning. The study aims to provide early warning, refined management, and dynamic strategy adjustments to prevent accidents and ensure safe coal mine production. The method involves collecting and analyzing data from 379 coal mine roof accidents, downscaling high-dimensional data using principal component analysis, and evaluating model performance using algorithms such as KNN, SVM, and DT.

Key Takeaways:

• The study proposes a practical method for predicting the risk level of coal mine roof accidents, which can provide risk early warning, refined management, and dynamic strategy adjustments for accident prevention.
• The method involves collecting and analyzing data from coal mine roof accidents, downscaling high-dimensional data using principal component analysis, and evaluating model performance using machine learning algorithms.
• The Random Forest integration algorithm is introduced to improve the evaluation and prediction of the model, resulting in a significant improvement in prediction accuracy to 0.94 and recall rate to 0.87.
• The method can also be applied to the risk level prediction of other coal mine accidents, assisting coal mine operators in checking safety issues and taking precautions.
• The study highlights the importance of using machine learning techniques to predict and prevent coal mine accidents, which can improve control effectiveness and ensure safe coal mine production.

Statistics:

• 379 coal mine roof accidents data were collected for the study.
• 305 cases were established after screening and filtering.
• The prediction accuracy of the model jumps to 0.94 after using the Random Forest integration algorithm.
• The recall rate and F1 score of the model achieve a significant improvement to 0.87 and 0.89, respectively.
• The study concludes that the method can be applied to other coal mine accidents, assisting coal mine operators in checking safety issues and taking precautions for 0.94 and 0.87 instances.

Sources:

• Research on predicting the risk level of coal mine roof accident based on machine learning. Scientific Reports, 2025;15(1):24028.
• Nature Publishing Group (www.nature.com/)
• Scientific Reports (www.nature.com/srep/)
• China University of Mining and Technology, School of Safety Engineering, Xuzhou, 221116, People's Republic of China