Figure - available from: Frontiers in Earth Science
This content is subject to copyright.
Source publication
With the sustainable development of the construction industry, recycled aggregate (RA) has been widely used in concrete preparation to reduce the environmental impact of construction waste. Compressive strength is an essential measure of the performance of recycled aggregate concrete (RAC). In order to understand the correspondence between relevant...
Similar publications
Reinforced concrete (RC) structures are used in a wide range of applications, including high-rise buildings, nuclear power plants, oil and gas platforms, bridges, and other infrastructure. However, over time, RC structures can be subject to deterioration and damage, particularly from exposure to weather and environmental conditions, heavy traffic l...
Institutions of higher learning have made persistent efforts to provide students with a high-quality education. Educational data mining (EDM) enables academic institutions to gain insight into student data in order to extract information for making predictions. COVID-19 represents the most catastrophic pandemic in human history. As a result of the...
![The R2\documentclass[12pt]{minimal} \usepackage{amsmath}...](publication/370899739/figure/fig2/AS:11431281163103692@1685498803709/The-R2documentclass12ptminimal-usepackageamsmath-usepackagewasysym_Q320.jpg)
![The R2\documentclass[12pt]{minimal} \usepackage{amsmath}...](publication/370899739/figure/fig4/AS:11431281163093285@1685498803958/The-R2documentclass12ptminimal-usepackageamsmath-usepackagewasysym_Q320.jpg)
![The R2\documentclass[12pt]{minimal} \usepackage{amsmath}...](publication/370899739/figure/fig5/AS:11431281163047011@1685498804240/The-R2documentclass12ptminimal-usepackageamsmath-usepackagewasysym_Q320.jpg)
The prediction accuracy of current mainstream machine learning (ML) models depends on regulating many hyperparameters. In this paper, a deep forest (DF) model with a few hyperparameters and a non-excessive dependence on super parameter regulation was applied to the prediction of glass-forming ability (GFA) of bulk metallic glasses (BMGs). Compared...
Machine Learning (ML) is a branch of Artificial Intelligence (AI) in which data-driven schemes learn patterns by being exposed to relevant data. Natural Language Processing (NLP) has been benefiting greatly from ML. In this article, ML is used to anticipate hotel review ratings. This paper proposes an Ensemble Learning-based rating prediction model...
A fire accident is the most tragic incident in human life. Particularly environmental hazards such as forest fires lead loss of wildlife, economy, wealth, human lives and pollution. our research purpose of predict the occurrence of fire incidents using ensemble machine learning models. The goal is to develop an accurate and reliable model that can...
Citations
... Although these parameters exhibited significant correlations, they were considered negligible in the context of CS prediction (Davawala et al. 2023;Peng and Unluer 2023). It can also be inferred that the adherence of mortar to the surface of RAC with higher water absorption is directly related to the RAC replacement rate (Duan and Poon 2014;Zhang et al. 2023). Furthermore, it was discovered that the inclusion of RCA amount, W, S, and WA_RA percentage negatively impacts RAC CS. ...
... This observation could be attributed to the RCA properties being closer to NCA. These findings align with previous studies (Peng and Unluer 2023;Zhang et al. 2023). However, it's important to emphasize that these specific outcomes are contingent upon the database utilized an RMSE of 4.045, it showed minimal forecast errors, highlighting its competence in approximating CS. ...
The construction industry’s shift towards sustainable practices has spurred interest in innovative materials, with Recycled Aggregate Concrete (RAC) standing out as a notable candidate. This material leverages recycled aggregates to mitigate waste, conserve resources, and reduce environmental impact. However, the accurate prediction of RAC’s compressive strength (CS) is challenging due to its intricate composition and variable material properties. To address this, artificial intelligence (AI) models are increasingly being used for their ability to uncover complex data patterns. This study offers a detailed comparison of ten advanced AI models for predicting RAC CS, including Artificial Neural Networks, Support Vector Regression, Decision Tree Regression, Random Forest Regression, k-Nearest Neighbors, Lasso, AdaBoost, Bagging, XGBoost, and CatBoost models. Each model is fine-tuned through hyperparameter optimization to enhance predictive accuracy. Additionally, SHAP (SHapley Additive exPlanations) algorithms are employed to interpret the models, providing insights into feature importance. The results demonstrate that all models achieved R² values exceeding 75%, with the CatBoost model attaining the highest R² value of 91% on the testing set. The CatBoost model also recorded the lowest error indices, with an MAE of 2.79 and an RMSE of 4.045, making it the most effective model for predicting RAC strength. SHAP analysis identified cement, water, sand, and RA water absorption as key features influencing RAC strength. This study underscores the potential of AI models in advancing the predictability and performance of sustainable construction materials.
... This method is repeated k times, using the validation data costly system. The surrogate model is adjusted repeatedly to include all currently visible objectives (Zhang et al. 2023). The effectiveness of various potential points is evaluated using the acquisition function, which is based on the prediction distribution of the probabilistic model. ...
Recycled aggregate concrete (RAC) has been adopted in building construction as it can reduce concrete waste, eventually minimizing the environmental impact. However, using recycled materials can lead to compromised performance of mechanical properties like split tensile strength (STS). Several factors, including density, water absorption, and recycled aggregate proportion, play a vital role in assessments of STS. This study explores the prediction of STS by using a hybridized machine learning algorithm in terms of performance indicators. Ensemble model XGB with five optimization algorithms, namely Random search (RS), Grid search (GS), Bayesian Optimization (BO), Grey Wolf optimization (GWO), and Particle Swarm Optimization (PSO), are considered for the study. The comparison shows that XGB-PSO performed very well with R² of 0.999 and 0.960 in the training and testing sets, respectively. The potential performance of GWO is also seen during the assessments. Further, the 10-fold cross-validation used in this study ensures that the models can predict better without overfitting. The model’s explainability is done using Shapley Additive Explanations (SHAP) analysis. SHAP-based study reveals that Cement, Water, and size of aggregates (M-RCA) are critical elements and may enhance STS if considered. The best cement range is 300 to 500 kg/m³, the M-RCA size is 10 to 20 mm, and the water range is 180 to 200 kg/m³. SHAP interaction graphs confirm the result. SHAP interaction graphs confirm a good agreement with the applied range of input parameters for the enhancement of STS with the experimental findings of earlier research.
... Yuan et al. [61] used ensemble ML methods to predict RAC's compressive and flexural strengths using twelve input factors from the dataset and analyzing their effect on RAC strength. Zhang et al. [63] developed a model to estimate RAC compressive strength using ML techniques and hyperparameter optimization. Specifically, The impact on model prediction efficiency and accuracy of different hyperparameter optimization techniques was investigated. ...
... Additionally, this study modeled more algorithms that have good accuracy, such as LightGBM, for this dataset, while the data pre-processing was not done and the data were entered into the modeling without removing the noise cases, and the models were able to achieve even better accuracy. Specifically, the MAPE achieved in our study, 5.46 MPa, surpassed that reported in Zhang et al. [63]'s work, which stood at 7.39 MPa. Moreover, this study has conducted more hyperparameter tuning methods, including successive halving, compared to Zhang et al. [63]'s study. ...
... Specifically, the MAPE achieved in our study, 5.46 MPa, surpassed that reported in Zhang et al. [63]'s work, which stood at 7.39 MPa. Moreover, this study has conducted more hyperparameter tuning methods, including successive halving, compared to Zhang et al. [63]'s study. While this study has introduced a quantitative index to compare the performance of models before applying hyperparameter tuning and after it, and the standard deviation of models has been added as a parameter to evaluate models, which has not been seen in Zhang et al. [63]'s study. ...
RAC is a kind of concrete made from Recycled Concrete Aggregates instead of natural aggregates. The use of RAC has been popular in recent years due to the environmental benefits of reducing waste and preserving natural resources. However, one of the RAC-using challenges is accurately predicting its compressive strength. This is a crucial factor in determining its suitability for various structural applications. In this research, eight ML algorithms were trained using a dataset of RAC samples to predict their compressive strength. They were Random Forest, support vector regression (SVR), K nearest neighbors (KNN), light gradient boosting machine (Light GBM), adaptive boosting (Adaboost), gradient boosting, extreme gradient boosting (XGBoost), and multi-layer perceptron (MLP). The best hyperparameters for each algorithm obtained using different hyperparameter tuning methods include Grid Search, Random Search, Successive Halving, Bayesian Optimization with Gaussian Processes (BOGP), Bayesian Optimization Random Forest (BORF), and Bayesian optimization gradient boosted trees (BOGB). The study’s findings indicated that Gradient Boosting has the highest performance in predicting the compressive strength of RAC after applying the hyperparameter tuning methods, with R2 and RMSE equal to 0.86 and 5.46 MPa, respectively. In addition, a sensitivity analysis was performed to determine the effect of various input parameters on the compressive strength of RAC. This indicated that the Effective Water-Cement ratio and the RAC Nominal maximum size had the most significant effect. The results show the potential of machine learning algorithms to predict the compressive strength of RAC, which can contribute to the development of more sustainable building materials.
... During training and prediction, multiple hyperparameters of the machine learning model must be configured [20][21][22][23], with the hyperparameter values closely linked to the prediction performance. In this regard, when performing hyperparameter adjustment and optimization for the aforementioned three algorithms: SVR, XGBoost, and ANN, the authors employed the Tree-structured Parzen Estimator (TPE) method for SVR and XGBoost, and the Grid Search method for ANN. ...
To investigate the impact of structural damages on the comfort level of suspension footbridges under human-induced vibrations, this study addresses the limitations of traditional manual testing, which often entails significant manpower and material resources. The aim is to achieve rapid estimation and health monitoring of comfort levels during bridge operation. To accomplish this, the study combines finite-element simulation results to establish a data-driven library and introduces three distinct machine learning algorithms. Through comparative analysis, a machine learning-based method is proposed for quick evaluation of bridge comfort levels. Focusing on the Yangjiadong Suspension Bridge, the study evaluates and researches the comfort level of the structure under the influence of human-induced vibrations. The findings revealed a relatively low base frequency and high flexibility. Additionally, when considering the mass of individuals, peak acceleration decreased. The predictive performance of the Artificial Neural Network (ANN) model was found to be superior when accounting for multi-parameter damages, yielding root mean square error (RMSE), mean absolute percentage error (MAPE), and R-squared (R2) values of 0.03, 0.02, and 0.98, respectively. Moreover, the error ratio of the generalization performance analysis was below 5%. Furthermore, the study identified a damage coefficient of 0.13 for the bridge’s main cable, hanger, and steel longitudinal beam. Under a crowd density of 0.5 people per square meter, the predicted peak acceleration was 1.098 m/s2, with a model error of less than 10% compared to the observed value of 1.004 m/s2. These results underscore the model’s effectiveness in swiftly evaluating bridge comfort levels, thereby offering valuable insights for the health monitoring of bridge comfort levels.
... The study employed the grid search method to simultaneously find the optimal CV values and hyperparameter combinations for each model [75,76]. The range of CV fold values was set between a minimum of 2 and a maximum of 10, and various predefined hyperparameter values for each model were explored to find the best combination. ...
... The statement regarding the importance of the "Fine Aggregate/Total Material" feature and its influence on compressive strength is consistent with previous studies (e.g., [88]). Similarly, the "Water/Cement" feature is identified as the second most important, aligning with the general understanding that reducing the water-cement ratio can improve the compressive strength of concrete [1,43,75,89]. The third-largest impact is attributed to the "Age" feature, suggesting that curing time or the age of the concrete influences compressive strength, which of course is well known. ...
... The statement regarding the importance of the "Fine Aggregate/Total Material" feature and its influence on compressive strength is consistent with previous studies (e.g., [88]). Similarly, the "Water/Cement" feature is identified as the second most important, aligning with the general understanding that reducing the watercement ratio can improve the compressive strength of concrete [1,43,75,89]. The third-largest impact is attributed to the "Age" feature, suggesting that curing time or the age of the concrete influences compressive strength, which of course is well known. ...
To reduce the environmental impact of concrete, recycled aggregates are of significant interest. Recycled concrete aggregate (RCA) presents a significant resource opportunity, although its performance as an aggregate in concrete is variable. This study presents a meta-analysis of the published literature to refine the understanding of how the moisture content of RCA, as well as other parameters, affects the compressive strength of concrete. Seven machine learning models were used to predict the compressive strength of concrete with RCA, including linear regression, support vector regression (SVR), and k-nearest neighbors (KNN) as single models, and decision tree, random forest, XGBoost, and LightGBM as ensemble models. The results of this study demonstrate that ensemble models, particularly the LightGBM model, exhibited superior prediction accuracy compared to single models. The LightGBM model yielded the highest prediction accuracy with R2 = 0.94, RMSE = 4.16 MPa, MAE = 3.03 MPa, and Delta RMSE = 1.4 MPa, making it the selected final model. The study, employing feature importance with LightGBM as the final model, identified age, water/cement ratio, and fine RCA aggregate content as key factors influencing compressive strength in concrete with RCA. In an interaction plot analysis using the final model, lowering the water–cement ratio consistently improved compressive strength, especially between 0.3 and 0.4, while increasing the fine RCA ratio decreased compressive strength, particularly in the range of 0.4 to 0.6. Additionally, it was found that maintaining moisture conditions of RCA typically between 0.0 and 0.8 was crucial for maximizing strength, whereas extreme moisture conditions, like fully saturated surface dry (SSD) state, negatively impacted strength.
... The TPE is used as a proxy model to more accurately represent the behavior of a complicated or computationally costly system. The surrogate model is adjusted repeatedly to include all currently visible objectives (Zhang et al., 2023). The effectiveness of various potential points is evaluated using the acquisition function, which is based on the prediction distribution of the probabilistic model. ...
Recycled aggregate-based concrete has been adopted in building construction as it can reduce concrete waste, eventually minimizing the environmental impact. However, using recycled materials can lead to compromised performance of mechanical properties like split tensile strength (STS). Several factors, including density, water absorption, and recycled aggregate proportion, play a vital role in assessments of STS. This study explores the better evaluation of STS using a hybridized machine learning algorithm. Ensemble model XGBoost with five optimization algorithms, namely Random search (RS), Grid search (GS), Bayesian Optimization (BO), Grey Wolf optimization (GWO), and Particle Swarm Optimization (PSO) are considered for the study. The comparison shows that XGB-PSO performed very well with R ² of 0.9988 and 0.9602 in the training and testing sets, respectively. The potential performance of GWO is also seen during the assessments. Further, the 10-fold cross-validation used in this study ensures that the models can predict better without overfitting. The model's explainability is done using Shapley Additive Explanations (SHAP) analysis. SHAP-based study reveals that Cement, Water, and size of aggregates (M-RCA) are critical elements and may enhance STS if considered. The best cement range is 300 to 500 kg/m ³ , the M-RCA size is 10 to 20 mm, and the water range is 180 to 200 kg/m ³ . SHAP interaction graphs confirm the result. This study helps engineers and researchers to understand the critical parameters for making informed decisions, thus promoting sustainable construction practices.
... Computers, their hardware and software, and the supported devices significantly contribute to diverse aspects of human life (Diker, 2022), and one such aspect is yield DL and ML approaches have also used Bayesian optimization algorithms for hyperparameter optimization. including Three-Way Decision-based Bayesian Deep Learning (TWDBDL) for uncertainty quantification (Abdar et al., 2021), scalable BO using Deep Neural Networks (Snoek et al., 2015), predicting the properties of recycled aggregate concrete through Bayesian optimization techniques (Zhang et al., 2023b), and (Al-Rawashdeh et al., 2023) conducted a comparative investigation employing Bayesian optimization to predict building damage grade caused by earthquakes across multiple machine learning algorithms. ...
Accurate pre-harvest crop yield estimation is vital for agricultural sustainability and economic stability. The existing yield estimating models exhibit deficiencies in insufficient examination of hyperparameters, lack of robustness, restricted transferability of meta-models, and uncertain generalizability when applied to agricultural data. This study presents a novel meta-knowledge-guided framework that leverages three diverse agricultural datasets and explores meta-knowledge transfer in frequent hyperparameter optimization scenarios. The framework’s approach involves base tasks using LightGBM and Bayesian Optimization, which automates hyperparameter optimization by eliminating the need for manual adjustments. Conducted rigorous experiments to analyze the meta-knowledge transformation of RGPE, SGPR, and TransBO algorithms, achieving impressive
values (0.8415, 0.9865, 0.9708) using rgpe_prf meta-knowledge transfer on diverse datasets. Furthermore, the framework yielded excellent results for mean squared error (MSE), mean absolute error (MAE), scaled MSE, and scaled MAE. These results emphasize the method’s significance, offering valuable insights for crop yield estimation, benefiting farmers and the agricultural sector.
... In recent years, in the fields of technology in general and construction technology in particular, artificial neural networks (ANNs) have been studied and applied to model nonlinear relationships between object parameters [22][23][24] as well as the behavior of materials. According to Haykin [25], the advantage of using ANN is the capability to solve problems that do not have a known or given mathematical model of the object. ...
This article presents a solution using an artificial neural network and a neuro-fuzzy network to predict the rate of water evaporation and the size of the shrinkage of a self-compacting concrete mixture based on the concrete mixture parameters and the environment parameters. The concrete samples were mixed and measured at four different environmental conditions (i.e., humid, dry, hot with high humidity, and hot with low humidity), and two curing styles for the self-compacting concrete were measured. Data were collected for each sample at the time of mixing and pouring and every 60 minutes for the next ten hours to help create prediction models for the required parameters. A total of 528 samples were collected to create the training and testing data sets. The study proposed to use the classic Multi-Layer Perceptron and the modified Takaga-Sugeno-Kang neuro-fuzzy network to estimate the water evaporation rate and the shrinkage size of the concrete sample when using four inputs: the concrete water-to-binder ratio, environment temperature, relative humidity, and the time after pouring the concrete into the mold. Real-field experiments and numerical computations have shown that both of the models are good as parameter predictors, where low errors can be achieved. Both proposed networks achieved for testing results R2 bigger than 0.98, the mean of squared errors for water evaporation percentage was less than 1.43%, and the mean of squared errors for shrinkage sizes was less than 0.105 mm/m. The computation requirements of the two models in testing mode are also low, which can allow their easy use in practical applications. Doi: 10.28991/CEJ-2024-010-01-07 Full Text: PDF
... Grid search and TPE optimization are widely used hyperparameter tuning techniques (Alibrahim and Ludwig, 2021). Some scholars comparing these two methods have found that TPE is superior in most cases Zhang et al., 2023). Hence, the TPE, a Bayesian-based parameter tuning strategy, is applied to optimize the hyperparameters of models. ...
Wheel and rail profiles have significant impacts on the vehicle system dynamics. Improper wheel and rail profiles lead to the decay of vehicle dynamics performance, hunting instability and derailment. In this study, we propose a model, stacking feature deep forest, to determine the limit wheel profile for hunting instability using wheel/rail geometric contact parameters. We calculate the critical speed of the vehicles under various wheel and rail profiles and record their wheel/rail geometric contact parameters as samples. Two experiments are conducted. One is the binary classification aiming to identify whether the wheel/rail geometric contact parameters guarantee the vehicle's critical speed is higher than the threshold of its operational speed. The other is the multi-classification to determine the interval of the critical speed. The results show the stacking feature deep forest achieves an accuracy of 85.10% on the test set which is higher than other popular ensemble models. However, all models' accuracy is lower than 70% in multi-classification. The Shapley additive explanations are utilized to enhance the explainability of the stacking feature deep forest. The Shapley additive explanation importance reveals that the equivalent conicity of 1 mm has the most impact on hunting. Its importance is 1.32 times more than the equivalent conicity of 3 mm. Nevertheless, the equivalent conicity of 3 mm is nearly positively correlated to the hunting probability and this characteristic is suitable as a qualitative criterion.
