Decision tree models for the estimation of geo-polymer concrete compressive strength.

The green concretes industry benefits from utilizing gel to replace parts of the cement in concretes. However, measuring the compressive strength of geo-polymer concretes (CSGPoC) needs a significant amount of work and expenditure. Therefore, the best idea is predicting CSGPoC with a high level of accuracy. To do this, the base learner and super learner machine learning models were proposed in this study to anticipate CSGPoC. The decision tree (DT) is applied as base learner, and the random forest and extreme gradient boosting (XGBoost) techniques are used as super learner system. In this regard, a database was provided involving 259 CSGPoC data samples, of which four-fifths of is considered for the training model and one-fifth is selected for the testing models. The values of fly ash, ground-granulated blast-furnace slag (GGBS), Na2SiO3, NaOH, fine aggregate, gravel 4/10 mm, gravel 10/20 mm, water/solids ratio, and NaOH molarity were considered as input of the models to estimate CSGPoC. To evaluate the reliability and performance of the decision tree (DT), XGBoost, and random forest (RF) models, 12 performance evaluation metrics were determined. Based on the obtained results, the highest degree of accuracy is achieved by the XGBoost model with mean absolute error (MAE) of 2.073, mean absolute percentage error (MAPE) of 5.547, Nash-Sutcliffe (NS) of 0.981, correlation coefficient (R) of 0.991, R2 of 0.982, root mean square error (RMSE) of 2.458, Willmott's index (WI) of 0.795, weighted mean absolute percentage error (WMAPE) of 0.046, Bias of 2.073, square index (SI) of 0.054, p of 0.027, mean relative error (MRE) of -0.014, and a20 of 0.983 for the training model and MAE of 2.06, MAPE of 6.553, NS of 0.985, R of 0.993, R2 of 0.986, RMSE of 2.307, WI of 0.818, WMAPE of 0.05, Bias of 2.06, SI of 0.056, p of 0.028, MRE of -0.015, and a20 of 0.949 for the testing model. By importing the testing set into trained models, values of 0.8969, 0.9857, and 0.9424 for R2 were obtained for DT, XGBoost, and RF, respectively, which show the superiority of the XGBoost model in CSGPoC estimation. In conclusion, the XGBoost model is capable of more accurately predicting CSGPoC than DT and RF models.

Assessment of the ground vibration during blasting in mining projects using different computational approaches.

The investigation compares the conventional, advanced machine, deep, and hybrid learning models to introduce an optimum computational model to assess the ground vibrations during blasting in mining projects. The long short-term memory (LSTM), artificial neural network (ANN), least square support vector machine (LSSVM), ensemble tree (ET), decision tree (DT), Gaussian process regression (GPR), support vector machine (SVM), and multilinear regression (MLR) models are employed using 162 data points. For the first time, the blackhole-optimized LSTM model has been used to predict the ground vibrations during blasting. Fifteen performance metrics have been implemented to measure the prediction capabilities of computational models. The study concludes that the blackhole optimized-LSTM model PPV11 is highly capable of predicting ground vibration. Model PPV11 has assessed ground vibrations with RMSE = 0.0181 mm/s, MAE = 0.0067 mm/s, R = 0.9951, a20 = 96.88, IOA = 0.9719, IOS = 0.0356 in testing. Furthermore, this study reveals that the prediction accuracy of hybrid models is less affected by multicollinearity because of the optimization algorithm. The external cross-validation and literature validation confirm the prediction capabilities of model PPV11. The ANOVA and Z tests reject the null hypothesis for actual ground vibration, and the Anderson-Darling test rejects the null hypothesis for predicted ground vibration. This study also concludes that the GPR and LSSVM models overfit because of moderate to problematic multicollinearity in assessing ground vibration during blasting.

Prediction of ground vibration due to mine blasting in a surface lead-zinc mine using machine learning ensemble techniques.

Ground vibration due to blasting is identified as a challenging issue in mining and civil activities. Peak particle velocity (PPV) is one of the blasting undesirable consequences, which is resulted during emission of vibration in blasted bench. This study focuses on the PPV prediction in the surface mines. In this regard, two ensemble systems, i.e., the ensemble of artificial neural networks and the ensemble of extreme gradient boosting (EXGBoosts) were developed for PPV prediction in one of the largest lead-zinc open-pit mines in the Middle East. For ensemble modeling, several ANN and XGBoost base models were separately designed with different architectures. Then, the validation indices such as coefficient determination (R2), root mean square error (RMSE), mean absolute error (MAE), the variance accounted for (VAF), and Accuracy were used to evaluate the performance of the base models. The five top base models with high accuracy were selected to construct an ensemble model for each of the methods, i.e., ANNs and XGBoosts. To combine the outputs of the top base models and achieve a single result stacked generalization technique, was employed. Findings showed ensemble models increase the accuracy of PPV predicting in comparison with the best individual models. The EXGBoosts was superior method for predicting of the PPV, which obtained values of R2, RMSE, MAE, VAF, and Accuracy corresponding to the EXGBoosts were (0.990, 0.391, 0.257, 99.013(%), 98.216), and (0.968, 0.295, 0.427, 96.674(%), 96.059), for training and testing datasets, respectively. However, the sensitivity analysis indicated that the spacing (r = 0.917) and number of blast-holes (r = 0.839) had the highest and lowest impact on the PPV intensity, respectively.