Statistical and Sensitivity Evaluation of Advanced Computational Models for Predicting Rock Mechanical Properties

Comparative Evaluation of Intelligent Machine Learning Models for Prediction of Uniaxial Compressive Strength of Rocks

Real-Time Prediction of Rock Young Modulus and Uniaxial Compressive Strength from Artificial Intelligence-Based Correlations

Rock geomechanical properties are vital parameters used for field development studies. Young's modulus (E) and uniaxial compressive strength (UCS) are two of the most fundamental variables used to characterize rock strength and formation deformation behavior. The conventional laboratory measurements of these parameters are costly, time-consuming and often not available during drilling operations. Consequently, real-time estimation of Static young modulus and UCS are crucial for drilling optimization, wellbore stability analysis and fracture containment assessment. This study investigates an artificial intelligence (AI) based approach for real-time prediction of Young's modulus and UCS using drilling and logging data. Ensemble supervised machine learning models were assessed to find the most accurate predictive framework. The Machine Learning models were developed using offset field dataset from five wells, comprising approximately 5,600 depth-indexed samples. The input variables includes rate of penetration (ROP), weight on bit (WOB), torque (T), Equivalent circulation density (ECD) and Gamma Ray (GR) and acoustic log responses of density and neutron. Four machine learning algorithms were investigated and compared, including Extreme gradient boosting (XGBoost), Random forest (RF), Extreme trees (ET) and Categorical Boosting (CB). The datasets were preprocessed, normalized and divided into training and testing subsets using 80:20 split respectively. During the model development, hyperparameter tuning was performed to optimize model performance and two independent wells not used during model development were used for blind testing to evaluate the generalization capability of the models. The results showed that all algorithms achieved a coefficient of determination (R2 >0.90) on the test datasets with ET and CB provided the best overall predictive performance. The optimized models achieved coefficients of determination (R2) of 0.88 for UCS and 0.875 for Young's modulus, with corresponding root mean square errors of 1484 psi and 3.59 GPa, respectively. Features importance analysis revealed that neutron porosity (NPHI), formation bulk density (RHOB) and Gamma Ray are the three most influential predictors. The proposed framework enables continuous, depth-based estimation of rock mechanical properties in real time, supporting improved geomechanical modeling and drilling optimization. The approach shows strong potential for integration into real-time drilling advisory systems for autonomous drilling workflows

Application of data-driven modeling techniques for predicting the shear strength of reinforced concrete beams

Accurate prediction of the shear strength of reinforced concrete (RC) beams remains a challenging problem due to the complex nonlinear interactions among material properties, reinforcement characteristics, and beam geometry. This study presents a data-driven artificial neural network (ANN) framework for predicting the shear strength of RC beams using a systematically curated experimental database comprising 1,977 specimens collected from published literature. The database was preprocessed to remove incomplete and duplicate records, and the optimal ANN architecture was selected using the total goodness function. Model performance was evaluated using 10-fold cross-validation together with multiple statistical metrics, including the coefficient of determination (R²), mean absolute error, root mean squared error, bias, and prediction interval. The ANN achieved R² values of 0.9968 and 0.9686 for the representative training and testing datasets, respectively, and an overall R² of 0.993, with 1,668 predictions falling within the ± 30% error criterion. Comparative evaluation with the Canadian Standards Association (CSA), American Concrete Institute 318 (ACI 318), and Eurocode 2 (EC2) design-code models demonstrated that the proposed ANN consistently achieved superior predictive accuracy and reliability. Parametric analyses further confirmed that the predicted trends agree with established reinforced concrete shear mechanics, highlighting the dominant influence of the shear span-to-depth ratio, beam geometry, and reinforcement ratio on shear resistance. The proposed framework provides an accurate and robust decision-support tool that complements conventional design-code methods for predicting the shear strength of RC beams.

Hybrid data-driven model for predicting the peak shear strength of rock joints

Sensitivity-Driven Evolutionary Polynomial Regression for Tropical Subgrade Permanent Deformation

This study developed an interpretable Evolutionary Polynomial Regression framework to predict permanent deformation of tropical soil subgrades in semi-rigid pavement structures. The database combined repeated-load triaxial test parameters reported in Brazilian studies with controlled mechanistic-empirical simulations representing traffic demand, structural thickness, subgrade Poisson’s ratio, and soil properties. Candidate equations were generated through a hybrid evolutionary search combining Genetic Algorithm and Differential Evolution, and final models were selected by jointly considering statistical performance, parsimony, and Monte Carlo sensitivity consistency. Six predictors were retained: number of axle-load repetitions, percentage passing the No. 200 sieve, optimum moisture content, clayeyness coefficient, laterization index, and equivalent pavement thickness. Model assessment used three repeated random train-test partitions. The selected equations contained three polynomial terms and achieved testing coefficient-of-determination values from 0.943 to 0.951, root mean square errors from 0.214 to 0.238 mm, and mean absolute errors from 0.163 to 0.169 mm. Sensitivity and Shapley analyses showed physically consistent trends, including increased deformation with traffic loading and laterization index, and reduced deformation with equivalent pavement thickness.

Machine Learning Models for Predicting Mechanical Properties of FRP-Confined Concrete Columns Across Low- to Ultra-High-Strength Concrete

This study presents a comprehensive analysis and predictive modeling framework for the axial compressive strength (fcc) and ultimate axial strain (εcu) of concrete columns confined within fiber-reinforced polymer (FRP) systems. Large databases comprising 3312 samples for fcc and 3319 for εcu were compiled from the literature, encompassing a wide range of key variables, including unconfined concrete strength from 7 MPa to 204 MPa and diverse FRP confinement configurations. The datasets were subjected to extensive statistical and multivariate analyses to identify the primary factors influencing axial behavior and guide feature selection for predictive modeling. Three groups of machine learning (ML) algorithms were subsequently considered: (i) artificial neural networks (including multilayer perceptrons with one and two hidden layers), (ii) kernel-based models (Gaussian process regression and support vector regression), and (iii) tree-based ensemble models (gradient boosting machine, eXtreme gradient boosting, and light gradient boosting machine). Hyperparameters were optimized using grid search cross-validation, while feature importance analyses were performed to quantify the contribution of each input variable. Among all ML models, eXtreme gradient boosting demonstrated superior predictive performance, effectively capturing the nonlinear and multivariate interactions governing confinement effectiveness. Comparative analysis with the top performing regression-based formulations further highlighted the accuracy, robustness, and generalization capability of the eXtreme gradient boosting model. The findings provide a data-driven and interpretable framework for the design and prediction of FRP-confined concrete columns.