Retrogressive landslides in sensitive clays pose significant risks to nearby infrastructure,as natural toe erosion or localized disturbances can trigger progressive block failures.While prior studies have largely reli...Retrogressive landslides in sensitive clays pose significant risks to nearby infrastructure,as natural toe erosion or localized disturbances can trigger progressive block failures.While prior studies have largely relied on two-dimensional(2D)large-deformation analyses,such models overlook key three-dimensional(3D)failure mechanisms and variability effects.This study develops a 3D probabilistic framework by integrating the Coupled Eulerian–Lagrangian(CEL)method with random field theory to simulate retrogressive landslides in spatially variable clay.Using Monte Carlo simulations,we compare 2D and 3D random large-deformation models to evaluate failure modes,runout distances,sliding velocities,and influence zones.The 3D analyses captured more complex failure modes—such as lateral retrogression and asynchronous block mobilization across slope width.Additionally,the 3D analyses predict longer mean runout distances(13.76 vs.11.92 m),wider mean influence distance(11.35 vs.8.73 m),and higher mean sliding velocities(4.66 vs.3.94 m/s)than their 2D counterparts.Moreover,3D models exhibit lower coefficients of variation(e.g.,0.10 for runout distance)due to spatial averaging across slope width.Probabilistic hazard assessment shows that 2D models significantly underpredict near-field failure probabilities(e.g.,48.8%vs.89.9%at 12 m from the slope toe).These findings highlight the limitations of 2D analyses and the importance of multi-directional spatial variability for robust geohazard assessments.The proposed 3D framework enables more realistic prediction of landslide mobility and supports the design of safer,risk-informed infrastructure.展开更多
Research on reconstructing imperfect faces is a challenging task.In this study,we explore a data-driven approach using a pre-trained MICA(MetrIC fAce)model combined with 3D printing to address this challenge.We propos...Research on reconstructing imperfect faces is a challenging task.In this study,we explore a data-driven approach using a pre-trained MICA(MetrIC fAce)model combined with 3D printing to address this challenge.We propose a training strategy that utilizes the pre-trained MICA model and self-supervised learning techniques to improve accuracy and reduce the time needed for 3D facial structure reconstruction.Our results demonstrate high accuracy,evaluated by the geometric loss function and various statistical measures.To showcase the effectiveness of the approach,we used 3D printing to create a model that covers facial wounds.The findings indicate that our method produces a model that fits well and achieves comprehensive 3D facial reconstruction.This technique has the potential to aid doctors in treating patients with facial injuries.展开更多
The main goal of this paper is to present the free vibration and buckling of viscoelastic functionally graded porous(FGP)nanosheet based on nonlocal strain gradient(NSGT)and surface elasticity theories.The nanosheets ...The main goal of this paper is to present the free vibration and buckling of viscoelastic functionally graded porous(FGP)nanosheet based on nonlocal strain gradient(NSGT)and surface elasticity theories.The nanosheets are placed on a visco-Pasternak medium in a hygro-temperature environment with nonlinear rules.The viscoelastic material characteristics of nanosheets are based on Kelvin’s model.The unique point of this study is to consider the change of nonlocal and length-scale coefficients according to thickness,similar to the laws of the material properties.The Galerkin approach based on the Kirchhoff-love plate theory is applied to determine the natural frequency and critical buckling load of the viscoelastic FGP nanosheet with various boundary conditions.The accuracy of the proposed method is verified through reliable publications.The outcome of this study highlights the significant effects of the nonlocal and length-scale parameters on the vibration and buckling behaviors of viscoelastic FGP nanosheets.展开更多
Efficiently executing inference tasks of deep neural networks on devices with limited resources poses a significant load in IoT systems.To alleviate the load,one innovative method is branching that adds extra layers w...Efficiently executing inference tasks of deep neural networks on devices with limited resources poses a significant load in IoT systems.To alleviate the load,one innovative method is branching that adds extra layers with classification exits to a pre-trained model,enabling inputs with high-confidence predictions to exit early,thus reducing inference cost.However,branching networks,not originally tailored for IoT environments,are susceptible to noisy and out-of-distribution(OOD)data,and they demand additional training for optimal performance.The authors introduce BrevisNet,a novel branching methodology designed for creating on-device branching models that are both resourceadaptive and noise-robust for IoT applications.The method leverages the refined uncertainty estimation capabilities of Dirichlet distributions for classification predictions,combined with the superior OOD detection of energy-based models.The authors propose a unique training approach and thresholding technique that enhances the precision of branch predictions,offering robustness against noise and OOD inputs.The findings demonstrate that BrevisNet surpasses existing branching techniques in training efficiency,accuracy,overall performance,and robustness.展开更多
In this paper,the isogeometric analysis(IGA)method is employed to analyze the oscillation characteristics of functionally graded triply periodic minimal surface(FG-TPMS)curved-doubly shells integrated with magneto-ele...In this paper,the isogeometric analysis(IGA)method is employed to analyze the oscillation characteristics of functionally graded triply periodic minimal surface(FG-TPMS)curved-doubly shells integrated with magneto-electric surface layers(referred to as"FG-TPMS-MEE curved-doubly shells")subjected to low-velocity impact loads.This study presents low-velocity impact load model based on a single springmass(S-M)approach.The FG-TPMS-MEE curved-doubly shells are covered with two magneto-electric surface layers,while the core layer consists of three types:I-graph and Wrapped Package-graph(IWP),Gyroid(G),and Primitive(P),with various graded functions.These types are notable for their exceptional stiffness-to-weight ratios,enabling a wide range of potential applications.The Maxwell equations and electromagnetic boundary conditions are applied to compute the change in electric potentials and magnetic potentials.The equilibrium equations of the shell are derived from a refined higher-order shear deformation theory(HSDT),and the transient responses of the FG-TPMS-MEE curveddoubly shells are subsequently determined using Newmark's direct integration method.These results have applications in structural vibration control and the analysis of structures subjected to impact or explosive loads.Furthermore,this study provides a theoretical prediction of the low-velocity impact load and magneto-electric-elastic effects on the free vibration and transient response of FG-TPMS-MEE curved-doubly shells.展开更多
Facial wound segmentation plays a crucial role in preoperative planning and optimizing patient outcomes in various medical applications.In this paper,we propose an efficient approach for automating 3D facial wound seg...Facial wound segmentation plays a crucial role in preoperative planning and optimizing patient outcomes in various medical applications.In this paper,we propose an efficient approach for automating 3D facial wound segmentation using a two-stream graph convolutional network.Our method leverages the Cir3D-FaIR dataset and addresses the challenge of data imbalance through extensive experimentation with different loss functions.To achieve accurate segmentation,we conducted thorough experiments and selected a high-performing model from the trainedmodels.The selectedmodel demonstrates exceptional segmentation performance for complex 3D facial wounds.Furthermore,based on the segmentation model,we propose an improved approach for extracting 3D facial wound fillers and compare it to the results of the previous study.Our method achieved a remarkable accuracy of 0.9999993% on the test suite,surpassing the performance of the previous method.From this result,we use 3D printing technology to illustrate the shape of the wound filling.The outcomes of this study have significant implications for physicians involved in preoperative planning and intervention design.By automating facial wound segmentation and improving the accuracy ofwound-filling extraction,our approach can assist in carefully assessing and optimizing interventions,leading to enhanced patient outcomes.Additionally,it contributes to advancing facial reconstruction techniques by utilizing machine learning and 3D bioprinting for printing skin tissue implants.Our source code is available at https://github.com/SIMOGroup/WoundFilling3D.展开更多
The task of mining erasable patterns(EPs)is a data mining problem that can help factory managers come up with the best product plans for the future.This problem has been studied by many scientists in recent times,and ...The task of mining erasable patterns(EPs)is a data mining problem that can help factory managers come up with the best product plans for the future.This problem has been studied by many scientists in recent times,and many approaches for mining EPs have been proposed.Erasable closed patterns(ECPs)are an abbreviated representation of EPs and can be con-sidered condensed representations of EPs without information loss.Current methods of mining ECPs identify huge numbers of such patterns,whereas intelligent systems only need a small number.A ranking process therefore needs to be applied prior to use,which causes a reduction in efficiency.To overcome this limitation,this study presents a robust method for mining top-rank-k ECPs in which the mining and ranking phases are combined into a single step.First,we propose a virtual-threshold-based pruning strategy to improve the mining speed.Based on this strategy and dPidset structure,we then develop a fast algorithm for mining top-rank-k ECPs,which we call TRK-ECP.Finally,we carry out experiments to compare the runtime of our TRK-ECP algorithm with two algorithms modified from dVM and TEPUS(Top-rank-k Erasable Pattern mining Using the Subsume concept),which are state-of-the-art algorithms for mining top-rank-k EPs.The results for the running time confirm that TRK-ECP outperforms the other experimental approaches in terms of mining the top-rank-k ECPs.展开更多
For the first time, the isogeometric analysis(IGA) approach is used to model and analyze free and forced vibrations of doubly-curved magneto-electro-elastic(MEE) composite shallow shell resting on the visco-Pasternak ...For the first time, the isogeometric analysis(IGA) approach is used to model and analyze free and forced vibrations of doubly-curved magneto-electro-elastic(MEE) composite shallow shell resting on the visco-Pasternak foundation in a hygro-temperature environment. The doubly-curved MEE shallow shell types include spherical shallow shell, cylindrical shallow shell, saddle shallow shell, and elliptical shallow shell subjected to blast load are investigated. The Maxwell equation and electromagnetic boundary conditions are used to determine the vary of the electric and magnetic potentials. The MEE shallow shell's equations of motion are derived from Hamilton's principle and refined higher-order shear theory. Then, the IGA method is used to derive the laws of natural frequencies and dynamic responses of the shell under various boundary conditions. The accuracy of the model and method is verified through reliable numerical comparisons. Aside from this, the impact of the input parameters on the free and forced vibration of the doubly-curved MEE shallow shell is examined in detail. These results may be useful in the design and manufacture of military structures such as warships, fighter aircraft, drones and missiles.展开更多
The use of machine learning to predict student employability is important in order to analyse a student’s capability to get a job.Based on the results of this type of analysis,university managers can improve the empl...The use of machine learning to predict student employability is important in order to analyse a student’s capability to get a job.Based on the results of this type of analysis,university managers can improve the employability of their students,which can help in attracting students in the future.In addition,learners can focus on the essential skills identified through this analysis during their studies,to increase their employability.An effectivemethod calledOPT-BAG(OPTimisation of BAGging classifiers)was therefore developed to model the problem of predicting the employability of students.This model can help predict the employability of students based on their competencies and can reveal weaknesses that need to be improved.First,we analyse the relationships between several variables and the outcome variable using a correlation heatmap for a student employability dataset.Next,a standard scaler function is applied in the preprocessing module to normalise the variables in the student employability dataset.The training set is then input to our model to identify the optimal parameters for the bagging classifier using a grid search cross-validation technique.Finally,the OPT-BAG model,based on a bagging classifier with optimal parameters found in the previous step,is trained on the training dataset to predict student employability.The empirical outcomes in terms of accuracy,precision,recall,and F1 indicate that the OPT-BAG approach outperforms other cutting-edge machine learning models in terms of predicting student employability.In this study,we also analyse the factors affecting the recruitment process of employers,and find that general appearance,mental alertness,and communication skills are the most important.This indicates that educational institutions should focus on these factors during the learning process to improve student employability.展开更多
Religion is one of the social entities that has had a significant impact on the pandemic.The study’s goals are to investigate the relationship between well-being and fear of COVID-19,as well as to test whether religi...Religion is one of the social entities that has had a significant impact on the pandemic.The study’s goals are to investigate the relationship between well-being and fear of COVID-19,as well as to test whether religious beliefs mediate the effect of wellbeing on fear of COVID-19.The sample comprised of 433 participants in Vietnam.Independent Sample t-Test,One-way ANOVA,mediation analysis were used to analyze the data.In the levels of well-being,individuals who engage in religious services daily have higher levels than those hardly and never attend,and people from the age of 18 to 30 have higher levels than individuals from 31 to above 60 years.In addition,people aged from 51 to above 60 have higher levels of religious beliefs than people aged from 18 to 50.Females experience more fear of COVID-19 compared to males.The latter illustrates that religious beliefs mediate the effect of well-being on fear of COVID-19.Social workers and clinicians must prioritize older adults and people with chronic diseases for early mental interventions,and they should be aware of the role of religion in psychological treatment integration.展开更多
Nowadays,air pollution is a big environmental problem in develop-ing countries.In this problem,particulate matter 2.5(PM2.5)in the air is an air pollutant.When its concentration in the air is high in developing countr...Nowadays,air pollution is a big environmental problem in develop-ing countries.In this problem,particulate matter 2.5(PM2.5)in the air is an air pollutant.When its concentration in the air is high in developing countries like Vietnam,it will harm everyone’s health.Accurate prediction of PM2.5 concentrations can help to make the correct decision in protecting the health of the citizen.This study develops a hybrid deep learning approach named PM25-CBL model for PM2.5 concentration prediction in Ho Chi Minh City,Vietnam.Firstly,this study analyzes the effects of variables on PM2.5 concentrations in Air Quality HCMC dataset.Only variables that affect the results will be selected for PM2.5 concentration prediction.Secondly,an efficient PM25-CBL model that integrates a convolutional neural network(CNN)andBidirectionalLongShort-TermMemory(Bi-LSTM)isdeveloped.This model consists of three following modules:CNN,Bi-LSTM,and Fully connected modules.Finally,this study conducts the experiment to compare the performance of our approach and several state-of-the-art deep learning models for time series prediction such as LSTM,Bi-LSTM,the combination of CNN and LSTM(CNN-LSTM),and ARIMA.The empirical results confirm that PM25-CBL model outperforms other methods for Air Quality HCMC dataset in terms of several metrics including Mean Squared Error(MSE),Root Mean Squared Error(RMSE),Mean Absolute Error(MAE),and Mean Absolute Percentage Error(MAPE).展开更多
In recent years,visual facial forgery has reached a level of sophistication that humans cannot identify fraud,which poses a significant threat to information security.A wide range of malicious applications have emerge...In recent years,visual facial forgery has reached a level of sophistication that humans cannot identify fraud,which poses a significant threat to information security.A wide range of malicious applications have emerged,such as deepfake,fake news,defamation or blackmailing of celebrities,impersonation of politicians in political warfare,and the spreading of rumours to attract views.As a result,a rich body of visual forensic techniques has been proposed in an attempt to stop this dangerous trend.However,there is no comprehensive,fair,and unified performance evaluation to enlighten the community on best performing methods.The authors present a systematic benchmark beyond traditional surveys that provides in-depth insights into facial forgery and facial forensics,grounding on robustness tests such as contrast,brightness,noise,resolution,missing information,and compression.The authors also provide a practical guideline of the benchmarking results,to determine the characteristics of the methods that serve as a comparative reference in this never-ending war between measures and countermeasures.The authors’source code is open to the public.展开更多
In recent years,the field of 3D printing has heavily relied on expert knowledge and complex trial-and-error procedures to determine appropriate printing parameters that meet desired consumption specifications.This stu...In recent years,the field of 3D printing has heavily relied on expert knowledge and complex trial-and-error procedures to determine appropriate printing parameters that meet desired consumption specifications.This study introduces a novel method for predicting 10 printing parameters based on 7 geometric features and 3 target consumption constraints(time,length,weight).Rather than using a traditional autoencoder model,we implement a variant that combines a reverse model with a forward-pretrained model.The forward model,pre-trained using XGBoost,predicts the 3 target consumption parameters from the 7 geometric features and 10 printing parameters.The reverse model then generates the 10 printing parameters from the 7 geometric features and the desired 3 consumption constraints.Through staged training and optimized loss function adjustments,our model achieves an R2 of 0.9567,demonstrating its precise predictive capabilities and potential to optimize the 3D printing process while reducing reliance on expert intervention.展开更多
We investigate a new numerical procedure based on a bubble-enriched finite element formulation in combination with the implicit backward Euler scheme for nonlinear analysis of strip footings and stability of slopes.Th...We investigate a new numerical procedure based on a bubble-enriched finite element formulation in combination with the implicit backward Euler scheme for nonlinear analysis of strip footings and stability of slopes.The soil body is modeled as a perfect plastic Mohr-Coulomb material.The displacement field is approximated by a 4-node quadrilateral element discretization enhanced with bubble modes.Collapse loads and failure mechanisms in cohesive frictional soil are determined by solving a few Newton-Raphson iterations.Numerical results of the present approach are verified by both analytical solutions and other numerical solutions available in the literature.展开更多
Identifying crack and predicting crack propagation are critical processes for the risk assessment of engineering structures.Most traditional approaches to crack modeling are faced with issues of high computational cos...Identifying crack and predicting crack propagation are critical processes for the risk assessment of engineering structures.Most traditional approaches to crack modeling are faced with issues of high computational costs and excessive computing time.To address this issue,we explore the potential of deep learning(DL)to increase the efficiency of crack detection and forecasting crack growth.However,there is no single algorithm that can fit all data sets well or can apply in all cases since specific tasks vary.In the paper,we present DL models for identifying cracks,especially on concrete surface images,and for predicting crack propagation.Firstly,SegNet and U-Net networks are used to identify concrete cracks.Stochastic gradient descent(SGD)and adaptive moment estimation(Adam)algorithms are applied to minimize loss function during iterations.Secondly,time series algorithms including gated recurrent unit(GRU)and long short-term memory(LSTM)are used to predict crack propagation.The experimental findings indicate that the U-Net is more robust and efficient than the SegNet for identifying crack segmentation and achieves the most outstanding results.For evaluation of crack propagation,GRU and LSTM are used as DL models and results show good agreement with the experimental data.展开更多
基金supported by the National Key Research and Development Program of China(No.2024YFC2815400)the European Commission(Nos.HORIZON MSCA-2024-PF-01 and 101200637)+2 种基金the Opening Fund of the State Key Laboratory of Water Resources Engineering and Management at Wuhan University(No.2024SGG07)the Shandong Provincial Natural Science Foundation(No.ZR2025MS647)the Sand Hazards and Opportunities for Resilience,Energy,and Sustainability(SHORES)Center,funded by Tamkeen under the NYUAD Research Institute Award CG013.
文摘Retrogressive landslides in sensitive clays pose significant risks to nearby infrastructure,as natural toe erosion or localized disturbances can trigger progressive block failures.While prior studies have largely relied on two-dimensional(2D)large-deformation analyses,such models overlook key three-dimensional(3D)failure mechanisms and variability effects.This study develops a 3D probabilistic framework by integrating the Coupled Eulerian–Lagrangian(CEL)method with random field theory to simulate retrogressive landslides in spatially variable clay.Using Monte Carlo simulations,we compare 2D and 3D random large-deformation models to evaluate failure modes,runout distances,sliding velocities,and influence zones.The 3D analyses captured more complex failure modes—such as lateral retrogression and asynchronous block mobilization across slope width.Additionally,the 3D analyses predict longer mean runout distances(13.76 vs.11.92 m),wider mean influence distance(11.35 vs.8.73 m),and higher mean sliding velocities(4.66 vs.3.94 m/s)than their 2D counterparts.Moreover,3D models exhibit lower coefficients of variation(e.g.,0.10 for runout distance)due to spatial averaging across slope width.Probabilistic hazard assessment shows that 2D models significantly underpredict near-field failure probabilities(e.g.,48.8%vs.89.9%at 12 m from the slope toe).These findings highlight the limitations of 2D analyses and the importance of multi-directional spatial variability for robust geohazard assessments.The proposed 3D framework enables more realistic prediction of landslide mobility and supports the design of safer,risk-informed infrastructure.
文摘Research on reconstructing imperfect faces is a challenging task.In this study,we explore a data-driven approach using a pre-trained MICA(MetrIC fAce)model combined with 3D printing to address this challenge.We propose a training strategy that utilizes the pre-trained MICA model and self-supervised learning techniques to improve accuracy and reduce the time needed for 3D facial structure reconstruction.Our results demonstrate high accuracy,evaluated by the geometric loss function and various statistical measures.To showcase the effectiveness of the approach,we used 3D printing to create a model that covers facial wounds.The findings indicate that our method produces a model that fits well and achieves comprehensive 3D facial reconstruction.This technique has the potential to aid doctors in treating patients with facial injuries.
文摘The main goal of this paper is to present the free vibration and buckling of viscoelastic functionally graded porous(FGP)nanosheet based on nonlocal strain gradient(NSGT)and surface elasticity theories.The nanosheets are placed on a visco-Pasternak medium in a hygro-temperature environment with nonlinear rules.The viscoelastic material characteristics of nanosheets are based on Kelvin’s model.The unique point of this study is to consider the change of nonlocal and length-scale coefficients according to thickness,similar to the laws of the material properties.The Galerkin approach based on the Kirchhoff-love plate theory is applied to determine the natural frequency and critical buckling load of the viscoelastic FGP nanosheet with various boundary conditions.The accuracy of the proposed method is verified through reliable publications.The outcome of this study highlights the significant effects of the nonlocal and length-scale parameters on the vibration and buckling behaviors of viscoelastic FGP nanosheets.
基金Australian Research Council,Grant/Award Numbers:DE200101465,DP240101108。
文摘Efficiently executing inference tasks of deep neural networks on devices with limited resources poses a significant load in IoT systems.To alleviate the load,one innovative method is branching that adds extra layers with classification exits to a pre-trained model,enabling inputs with high-confidence predictions to exit early,thus reducing inference cost.However,branching networks,not originally tailored for IoT environments,are susceptible to noisy and out-of-distribution(OOD)data,and they demand additional training for optimal performance.The authors introduce BrevisNet,a novel branching methodology designed for creating on-device branching models that are both resourceadaptive and noise-robust for IoT applications.The method leverages the refined uncertainty estimation capabilities of Dirichlet distributions for classification predictions,combined with the superior OOD detection of energy-based models.The authors propose a unique training approach and thresholding technique that enhances the precision of branch predictions,offering robustness against noise and OOD inputs.The findings demonstrate that BrevisNet surpasses existing branching techniques in training efficiency,accuracy,overall performance,and robustness.
文摘In this paper,the isogeometric analysis(IGA)method is employed to analyze the oscillation characteristics of functionally graded triply periodic minimal surface(FG-TPMS)curved-doubly shells integrated with magneto-electric surface layers(referred to as"FG-TPMS-MEE curved-doubly shells")subjected to low-velocity impact loads.This study presents low-velocity impact load model based on a single springmass(S-M)approach.The FG-TPMS-MEE curved-doubly shells are covered with two magneto-electric surface layers,while the core layer consists of three types:I-graph and Wrapped Package-graph(IWP),Gyroid(G),and Primitive(P),with various graded functions.These types are notable for their exceptional stiffness-to-weight ratios,enabling a wide range of potential applications.The Maxwell equations and electromagnetic boundary conditions are applied to compute the change in electric potentials and magnetic potentials.The equilibrium equations of the shell are derived from a refined higher-order shear deformation theory(HSDT),and the transient responses of the FG-TPMS-MEE curveddoubly shells are subsequently determined using Newmark's direct integration method.These results have applications in structural vibration control and the analysis of structures subjected to impact or explosive loads.Furthermore,this study provides a theoretical prediction of the low-velocity impact load and magneto-electric-elastic effects on the free vibration and transient response of FG-TPMS-MEE curved-doubly shells.
文摘Facial wound segmentation plays a crucial role in preoperative planning and optimizing patient outcomes in various medical applications.In this paper,we propose an efficient approach for automating 3D facial wound segmentation using a two-stream graph convolutional network.Our method leverages the Cir3D-FaIR dataset and addresses the challenge of data imbalance through extensive experimentation with different loss functions.To achieve accurate segmentation,we conducted thorough experiments and selected a high-performing model from the trainedmodels.The selectedmodel demonstrates exceptional segmentation performance for complex 3D facial wounds.Furthermore,based on the segmentation model,we propose an improved approach for extracting 3D facial wound fillers and compare it to the results of the previous study.Our method achieved a remarkable accuracy of 0.9999993% on the test suite,surpassing the performance of the previous method.From this result,we use 3D printing technology to illustrate the shape of the wound filling.The outcomes of this study have significant implications for physicians involved in preoperative planning and intervention design.By automating facial wound segmentation and improving the accuracy ofwound-filling extraction,our approach can assist in carefully assessing and optimizing interventions,leading to enhanced patient outcomes.Additionally,it contributes to advancing facial reconstruction techniques by utilizing machine learning and 3D bioprinting for printing skin tissue implants.Our source code is available at https://github.com/SIMOGroup/WoundFilling3D.
文摘The task of mining erasable patterns(EPs)is a data mining problem that can help factory managers come up with the best product plans for the future.This problem has been studied by many scientists in recent times,and many approaches for mining EPs have been proposed.Erasable closed patterns(ECPs)are an abbreviated representation of EPs and can be con-sidered condensed representations of EPs without information loss.Current methods of mining ECPs identify huge numbers of such patterns,whereas intelligent systems only need a small number.A ranking process therefore needs to be applied prior to use,which causes a reduction in efficiency.To overcome this limitation,this study presents a robust method for mining top-rank-k ECPs in which the mining and ranking phases are combined into a single step.First,we propose a virtual-threshold-based pruning strategy to improve the mining speed.Based on this strategy and dPidset structure,we then develop a fast algorithm for mining top-rank-k ECPs,which we call TRK-ECP.Finally,we carry out experiments to compare the runtime of our TRK-ECP algorithm with two algorithms modified from dVM and TEPUS(Top-rank-k Erasable Pattern mining Using the Subsume concept),which are state-of-the-art algorithms for mining top-rank-k EPs.The results for the running time confirm that TRK-ECP outperforms the other experimental approaches in terms of mining the top-rank-k ECPs.
文摘For the first time, the isogeometric analysis(IGA) approach is used to model and analyze free and forced vibrations of doubly-curved magneto-electro-elastic(MEE) composite shallow shell resting on the visco-Pasternak foundation in a hygro-temperature environment. The doubly-curved MEE shallow shell types include spherical shallow shell, cylindrical shallow shell, saddle shallow shell, and elliptical shallow shell subjected to blast load are investigated. The Maxwell equation and electromagnetic boundary conditions are used to determine the vary of the electric and magnetic potentials. The MEE shallow shell's equations of motion are derived from Hamilton's principle and refined higher-order shear theory. Then, the IGA method is used to derive the laws of natural frequencies and dynamic responses of the shell under various boundary conditions. The accuracy of the model and method is verified through reliable numerical comparisons. Aside from this, the impact of the input parameters on the free and forced vibration of the doubly-curved MEE shallow shell is examined in detail. These results may be useful in the design and manufacture of military structures such as warships, fighter aircraft, drones and missiles.
文摘The use of machine learning to predict student employability is important in order to analyse a student’s capability to get a job.Based on the results of this type of analysis,university managers can improve the employability of their students,which can help in attracting students in the future.In addition,learners can focus on the essential skills identified through this analysis during their studies,to increase their employability.An effectivemethod calledOPT-BAG(OPTimisation of BAGging classifiers)was therefore developed to model the problem of predicting the employability of students.This model can help predict the employability of students based on their competencies and can reveal weaknesses that need to be improved.First,we analyse the relationships between several variables and the outcome variable using a correlation heatmap for a student employability dataset.Next,a standard scaler function is applied in the preprocessing module to normalise the variables in the student employability dataset.The training set is then input to our model to identify the optimal parameters for the bagging classifier using a grid search cross-validation technique.Finally,the OPT-BAG model,based on a bagging classifier with optimal parameters found in the previous step,is trained on the training dataset to predict student employability.The empirical outcomes in terms of accuracy,precision,recall,and F1 indicate that the OPT-BAG approach outperforms other cutting-edge machine learning models in terms of predicting student employability.In this study,we also analyse the factors affecting the recruitment process of employers,and find that general appearance,mental alertness,and communication skills are the most important.This indicates that educational institutions should focus on these factors during the learning process to improve student employability.
文摘Religion is one of the social entities that has had a significant impact on the pandemic.The study’s goals are to investigate the relationship between well-being and fear of COVID-19,as well as to test whether religious beliefs mediate the effect of wellbeing on fear of COVID-19.The sample comprised of 433 participants in Vietnam.Independent Sample t-Test,One-way ANOVA,mediation analysis were used to analyze the data.In the levels of well-being,individuals who engage in religious services daily have higher levels than those hardly and never attend,and people from the age of 18 to 30 have higher levels than individuals from 31 to above 60 years.In addition,people aged from 51 to above 60 have higher levels of religious beliefs than people aged from 18 to 50.Females experience more fear of COVID-19 compared to males.The latter illustrates that religious beliefs mediate the effect of well-being on fear of COVID-19.Social workers and clinicians must prioritize older adults and people with chronic diseases for early mental interventions,and they should be aware of the role of religion in psychological treatment integration.
文摘Nowadays,air pollution is a big environmental problem in develop-ing countries.In this problem,particulate matter 2.5(PM2.5)in the air is an air pollutant.When its concentration in the air is high in developing countries like Vietnam,it will harm everyone’s health.Accurate prediction of PM2.5 concentrations can help to make the correct decision in protecting the health of the citizen.This study develops a hybrid deep learning approach named PM25-CBL model for PM2.5 concentration prediction in Ho Chi Minh City,Vietnam.Firstly,this study analyzes the effects of variables on PM2.5 concentrations in Air Quality HCMC dataset.Only variables that affect the results will be selected for PM2.5 concentration prediction.Secondly,an efficient PM25-CBL model that integrates a convolutional neural network(CNN)andBidirectionalLongShort-TermMemory(Bi-LSTM)isdeveloped.This model consists of three following modules:CNN,Bi-LSTM,and Fully connected modules.Finally,this study conducts the experiment to compare the performance of our approach and several state-of-the-art deep learning models for time series prediction such as LSTM,Bi-LSTM,the combination of CNN and LSTM(CNN-LSTM),and ARIMA.The empirical results confirm that PM25-CBL model outperforms other methods for Air Quality HCMC dataset in terms of several metrics including Mean Squared Error(MSE),Root Mean Squared Error(RMSE),Mean Absolute Error(MAE),and Mean Absolute Percentage Error(MAPE).
基金QuỹĐổi mới sáng tạo Vingroup,Grant/Award Number:VINIF.2020.ThS.BK.10。
文摘In recent years,visual facial forgery has reached a level of sophistication that humans cannot identify fraud,which poses a significant threat to information security.A wide range of malicious applications have emerged,such as deepfake,fake news,defamation or blackmailing of celebrities,impersonation of politicians in political warfare,and the spreading of rumours to attract views.As a result,a rich body of visual forensic techniques has been proposed in an attempt to stop this dangerous trend.However,there is no comprehensive,fair,and unified performance evaluation to enlighten the community on best performing methods.The authors present a systematic benchmark beyond traditional surveys that provides in-depth insights into facial forgery and facial forensics,grounding on robustness tests such as contrast,brightness,noise,resolution,missing information,and compression.The authors also provide a practical guideline of the benchmarking results,to determine the characteristics of the methods that serve as a comparative reference in this never-ending war between measures and countermeasures.The authors’source code is open to the public.
基金funding by the Deutsche Forschungsgemeinschaft under Germany's Excellence Strategy within the Cluster of Excellence PhoenixD(EXC 2122,Project ID 390833453).
文摘In recent years,the field of 3D printing has heavily relied on expert knowledge and complex trial-and-error procedures to determine appropriate printing parameters that meet desired consumption specifications.This study introduces a novel method for predicting 10 printing parameters based on 7 geometric features and 3 target consumption constraints(time,length,weight).Rather than using a traditional autoencoder model,we implement a variant that combines a reverse model with a forward-pretrained model.The forward model,pre-trained using XGBoost,predicts the 3 target consumption parameters from the 7 geometric features and 10 printing parameters.The reverse model then generates the 10 printing parameters from the 7 geometric features and the desired 3 consumption constraints.Through staged training and optimized loss function adjustments,our model achieves an R2 of 0.9567,demonstrating its precise predictive capabilities and potential to optimize the 3D printing process while reducing reliance on expert intervention.
文摘We investigate a new numerical procedure based on a bubble-enriched finite element formulation in combination with the implicit backward Euler scheme for nonlinear analysis of strip footings and stability of slopes.The soil body is modeled as a perfect plastic Mohr-Coulomb material.The displacement field is approximated by a 4-node quadrilateral element discretization enhanced with bubble modes.Collapse loads and failure mechanisms in cohesive frictional soil are determined by solving a few Newton-Raphson iterations.Numerical results of the present approach are verified by both analytical solutions and other numerical solutions available in the literature.
基金The first author would like to thank European Commission H2020-MSCA-RISE BESTOFRAC project for research funding.
文摘Identifying crack and predicting crack propagation are critical processes for the risk assessment of engineering structures.Most traditional approaches to crack modeling are faced with issues of high computational costs and excessive computing time.To address this issue,we explore the potential of deep learning(DL)to increase the efficiency of crack detection and forecasting crack growth.However,there is no single algorithm that can fit all data sets well or can apply in all cases since specific tasks vary.In the paper,we present DL models for identifying cracks,especially on concrete surface images,and for predicting crack propagation.Firstly,SegNet and U-Net networks are used to identify concrete cracks.Stochastic gradient descent(SGD)and adaptive moment estimation(Adam)algorithms are applied to minimize loss function during iterations.Secondly,time series algorithms including gated recurrent unit(GRU)and long short-term memory(LSTM)are used to predict crack propagation.The experimental findings indicate that the U-Net is more robust and efficient than the SegNet for identifying crack segmentation and achieves the most outstanding results.For evaluation of crack propagation,GRU and LSTM are used as DL models and results show good agreement with the experimental data.
