Prediction of undrained shear strength based on KMR model

The accurate prediction of undrained shear strength is of significant importance in areas such as slope stability, earthquake resistance, and pile foundation design. Therefore, enhancing the accuracy of undrained shear strength prediction is crucial. The research results indicate that employing the KMR model can improve the computational accuracy by 4.4% to 18.9%. Furthermore, compared to conventional empirical formulas, the KMR model evidently processes data more rapidly and can predict relevant parameters more accurately. This method provides a new research idea to some extent for addressing the issue of low prediction accuracy of traditional machine learning models for geotechnical parameters.