Assessment of Machine Learning Models for the Prediction of Rate-Dependent Compressive Strength of Rocks

Abstract

The prediction of rate-dependent compressive strength of rocks in dynamic compression experiments is still a notable challenge. Four machine learning models were introduced and employed on a dataset of 164 experiments to achieve an accurate prediction of the rate-dependent compressive strength of rocks. Then, the relative importance of the seven input features was analyzed. The results showed that compared with the extreme learning machine (ELM), random forest (RF), and the original support vector regression (SVR) models, the correlation coefficient R2 of prediction results with the hybrid model that combines the particle swarm optimization (PSO) algorithm and SVR was highest in both the training set and the test set, both exceeding 0.98. The PSO-SVR model obtained a higher prediction accuracy and a smaller prediction error than the other three models in terms of evaluation metrics, which showed the possibility of the model as a rate-dependent compressive strength prediction tool. Additionally, besides the static compressive strength, the stress rate is the most important influence factor on the rate-dependent compressive strength of the rock among the listed input parameters. Moreover, the strain rate has a positive effect on the rock strength.

Appraisal of Different Artificial Intelligence Techniques for the Prediction of Marble Strength

“…Among other widely accepted forms of AI computing, RF observes a singular association between model embodiment and the predictive accuracy [64]. The Random Forest algorithm may be expressed as [67]:…”

“…Support Vector Machines are capable of solving classification and complicated nonlinear regression issues. The basic goal when applied to regression problems is to create a perfect classification surface that reduces the error in all training samples obtained from that surface [67].…”

Human Endocrine-Disrupting Effects of Phthalate Esters through Adverse Outcome Pathways: A Comprehensive Mechanism Analysis

“…The RF model is based on the Bagging method, each classification is independent, and different input data have negligible effects on the model [49]. The R 2 of the RF model and the XGboost model for the harmful outcomes related to the ED effects of PAEs were 0.751 and 0.948, respectively (R 2 is the percentage of the predictive variable that can explain the variation of the outcome variable; R 2 values closer to 1 indicate that the model can better fit the data [50]). To summarize, the performance of the XGboost model was better and more accurate than that of the RF model, which was similar to the findings of a study by Hong, et al [51], who showed that the XGboost model had the highest discriminant performance while predicting the severity of COVID-19 pneumonia.…”

Several machine learning techniques comparison for the prediction of the uniaxial compressive strength of carbonate rocks

“…In recent years, various modelling techniques, such as simple and multivariable regression analyses, fuzzy inference system, neural network, other machine-learning algorithms, etc., have gained more attention and perceived as the best models to be used the prediction of the strength of rock materials. A reliable database for the mechanical and the physical rock properties is critical for the predictive models to assess the UCS of the rock 3,13,[20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37] .…”

“…They reported that multiple regression model with the explanatory variables such as porosity and P-wave velocity was the best model to estimate the UCS of travertine. Yang et al 37 foreseen the rate-dependent compressive strength of rocks utilizing 164 experimental datasets by employing four machine learning models; extreme learning machine (ELM), random forest (RF), SVR and hybrid model of particular swarm optimization (PSO)-SVR. Their study showed that the PSO-SVR model provided a higher accuracy of prediction with a less prediction error compared to the other three models.…”