Discussion on “Prediction of Engineering Characteristics of Rock Masses Using Actual TBM Performance Data with Supervised and Unsupervised Learning Algorithms (a Case Study in Strong to Very Strong Igneous and Pyroclastic Rocks)” [Rock Mech Rock Eng 57:7223–7252]
Ghorbani E.
2025Springer
Rock Mechanics and Rock Engineering
2025
The published article titled “Prediction of Engineering Characteristics of Rock Masses Using Actual TBM Performance Data with Supervised and Unsupervised Learning Algorithms (a Case Study in Strong to Very Strong Igneous and Pyroclastic Rocks)” has been meticulously read and examined. This publication tried to employ different statistical and machine learning (ML) (least squares-support vector machine—LS-SVM)-based methods to predict rock mass uniaxial compressive strength (UCS) and rock quality designation (RQD) using tunnel boring machine (TBM)-driven operational data. However, there are several issues in the article regarding the input parameters, modeling process and methodology, and findings. Meanwhile, the two most critical issues are (1) the high collinearity and multicollinearity of the input parameters, which were raised due to the use of some input parameters calculated with other input features, and (2) the complete failure of using classification-based modeling in the published article for continuous variables. This issue stems from the manual discretization of the data, the use of incorrect and uninterpretable confusion matrices, and the lack of performance evaluations for the classification-based models. This discussion aims to examine and address these issues.
Machine learning , Rock mass , Rock mechanics , RQD , TBM , Tunneling , UCS
Text of the article Перейти на текст статьи
Department of Mining Engineering, School of Mining and Geosciences, Nazarbayev University, Astana, 010000, Kazakhstan
Department of Mining Engineering
10 лет помогаем публиковать статьи Международный издатель
Книга Публикация научной статьи Волощук 2026 Book Publication of a scientific article 2026