GEOPRO

RMR PREDICTION USING RF

Random Forest (RF) is a powerful ensemble machine learning algorithm used for predicting Rock Mass Rating (RMR) based on various geotechnical input parameters. It operates by building multiple decision trees during training and averaging their outputs to improve prediction accuracy and reduce overfitting. Inputs for RMR prediction typically include parameters like uniaxial compressive strength (UCS), rock quality designation (RQD), joint spacing, joint condition, and groundwater conditions. RF models can handle both nonlinear relationships and missing data, making them well-suited for field-based datasets. They also provide feature importance scores, helping engineers identify which geotechnical parameters most influence RMR. Compared to traditional empirical methods, RF offers higher flexibility and often improved predictive performance. As a result, RF-based models are increasingly used in preliminary site investigations and rock mass classification tasks.