In recent years,landslide susceptibility mapping has substantially improved with advances in machine learning.However,there are still challenges remain in landslide mapping due to the availability of limited inventory...In recent years,landslide susceptibility mapping has substantially improved with advances in machine learning.However,there are still challenges remain in landslide mapping due to the availability of limited inventory data.In this paper,a novel method that improves the performance of machine learning techniques is presented.The proposed method creates synthetic inventory data using Generative Adversarial Networks(GANs)for improving the prediction of landslides.In this research,landslide inventory data of 156 landslide locations were identified in Cameron Highlands,Malaysia,taken from previous projects the authors worked on.Elevation,slope,aspect,plan curvature,profile curvature,total curvature,lithology,land use and land cover(LULC),distance to the road,distance to the river,stream power index(SPI),sediment transport index(STI),terrain roughness index(TRI),topographic wetness index(TWI)and vegetation density are geo-environmental factors considered in this study based on suggestions from previous works on Cameron Highlands.To show the capability of GANs in improving landslide prediction models,this study tests the proposed GAN model with benchmark models namely Artificial Neural Network(ANN),Support Vector Machine(SVM),Decision Trees(DT),Random Forest(RF)and Bagging ensemble models with ANN and SVM models.These models were validated using the area under the receiver operating characteristic curve(AUROC).The DT,RF,SVM,ANN and Bagging ensemble could achieve the AUROC values of(0.90,0.94,0.86,0.69 and 0.82)for the training;and the AUROC of(0.76,0.81,0.85,0.72 and 0.75)for the test,subsequently.When using additional samples,the same models achieved the AUROC values of(0.92,0.94,0.88,0.75 and 0.84)for the training and(0.78,0.82,0.82,0.78 and 0.80)for the test,respectively.Using the additional samples improved the test accuracy of all the models except SVM.As a result,in data-scarce environments,this research showed that utilizing GANs to generate supplementary samples is promising because it can improve the predictive展开更多
Contamination events in water distribution networks(WDNs)can have a huge impact on water supply and public health;increasingly,online water quality sensors are deployed for real-time detection of contamination events....Contamination events in water distribution networks(WDNs)can have a huge impact on water supply and public health;increasingly,online water quality sensors are deployed for real-time detection of contamination events.Machine learning has been used to integrate multivariate time series water quality data at multiple stations for contamination detection;however,accurate extraction of spatial features in water quality signals remains challenging.This study proposed a contamination detection method based on generative adversarial networks(GANs).The GAN model was constructed to simultaneously consider the spatial correlation between sensor locations and temporal information of water quality indicators.The model consists of two networksda generator and a discriminatordthe outputs of which are used to measure the degree of abnormality of water quality data at each time step,referred to as the anomaly score.Bayesian sequential analysis is used to update the likelihood of event occurrence based on the anomaly scores.Alarms are then generated from the fusion of single-site and multi-site models.The proposed method was tested on a WDN for various contamination events with different characteristics.Results showed high detection performance by the proposed GAN method compared with the minimum volume ellipsoid benchmark method for various contamination amplitudes.Additionally,the GAN method achieved high accuracy for various contamination events with different amplitudes and numbers of anomalous water quality parameters,and water quality data from different sensor stations,highlighting its robustness and potential for practical application to real-time contamination events.展开更多
The classification of lung nodules is a challenging problem as the visual analysis of the nodules and non-nodules revealed homogenous textural patterns.In this work,an Auxiliary Classifier(AC)-Generative Adversarial Net...The classification of lung nodules is a challenging problem as the visual analysis of the nodules and non-nodules revealed homogenous textural patterns.In this work,an Auxiliary Classifier(AC)-Generative Adversarial Network(GAN)based Lung Cancer Classification(LCC)system is developed.The pro-posed AC-GAN-LCC system consists of three modules;preprocessing,Lungs Region Detection(LRD),and AC-GAN classification.A Wienerfilter is employed in the preprocessing module to remove the Gaussian noise.In the LRD module,only the lung regions(left and right lungs)are detected using itera-tive thresholding and morphological operations.In order to extract the lung region only,floodfilling and background subtraction.The detected lung regions are fed to the AC-GAN classifier to detect the nodules.It classifies the nodules into one of the two classes,i.e.,binary classification(such as nodules or non-nodules).The AC-GAN is the extended version of the conditional GAN that predicts the label of a given image.Three different optimization techniques,adaptive gradient optimi-zation,root mean square propagation optimization,and Adam optimization are employed for optimizing the AC-GAN architecture.The proposed AC-GAN-LCC system is evaluated on the Lung Image Database Consortium(LIDC)data-base Computed Tomography(CT)scan images.The proposed AC-GAN-LCC system classifies∼15000 CT slices(7310 non-nodules and 7685 nodules).It pro-vides an overall accuracy of 98.8%on the LIDC database using Adam optimiza-tion by a 10-fold cross-validation approach.展开更多
Recently developed fault classification methods for industrial processes are mainly data-driven.Notably,models based on deep neural networks have significantly improved fault classification accuracy owing to the inclu...Recently developed fault classification methods for industrial processes are mainly data-driven.Notably,models based on deep neural networks have significantly improved fault classification accuracy owing to the inclusion of a large number of data patterns.However,these data-driven models are vulnerable to adversarial attacks;thus,small perturbations on the samples can cause the models to provide incorrect fault predictions.Several recent studies have demonstrated the vulnerability of machine learning methods and the existence of adversarial samples.This paper proposes a black-box attack method with an extreme constraint for a safe-critical industrial fault classification system:Only one variable can be perturbed to craft adversarial samples.Moreover,to hide the adversarial samples in the visualization space,a Jacobian matrix is used to guide the perturbed variable selection,making the adversarial samples in the dimensional reduction space invisible to the human eye.Using the one-variable attack(OVA)method,we explore the vulnerability of industrial variables and fault types,which can help understand the geometric characteristics of fault classification systems.Based on the attack method,a corresponding adversarial training defense method is also proposed,which efficiently defends against an OVA and improves the prediction accuracy of the classifiers.In experiments,the proposed method was tested on two datasets from the Tennessee–Eastman process(TEP)and steel plates(SP).We explore the vulnerability and correlation within variables and faults and verify the effectiveness of OVAs and defenses for various classifiers and datasets.For industrial fault classification systems,the attack success rate of our method is close to(on TEP)or even higher than(on SP)the current most effective first-order white-box attack method,which requires perturbation of all variables.展开更多
基金This research is funded by the Centre for Advanced Modeling and Geospatial Information Systems(CAMGIS),Faculty of Engineering and Information Technology,the University of Technology Sydney,Australia.
文摘In recent years,landslide susceptibility mapping has substantially improved with advances in machine learning.However,there are still challenges remain in landslide mapping due to the availability of limited inventory data.In this paper,a novel method that improves the performance of machine learning techniques is presented.The proposed method creates synthetic inventory data using Generative Adversarial Networks(GANs)for improving the prediction of landslides.In this research,landslide inventory data of 156 landslide locations were identified in Cameron Highlands,Malaysia,taken from previous projects the authors worked on.Elevation,slope,aspect,plan curvature,profile curvature,total curvature,lithology,land use and land cover(LULC),distance to the road,distance to the river,stream power index(SPI),sediment transport index(STI),terrain roughness index(TRI),topographic wetness index(TWI)and vegetation density are geo-environmental factors considered in this study based on suggestions from previous works on Cameron Highlands.To show the capability of GANs in improving landslide prediction models,this study tests the proposed GAN model with benchmark models namely Artificial Neural Network(ANN),Support Vector Machine(SVM),Decision Trees(DT),Random Forest(RF)and Bagging ensemble models with ANN and SVM models.These models were validated using the area under the receiver operating characteristic curve(AUROC).The DT,RF,SVM,ANN and Bagging ensemble could achieve the AUROC values of(0.90,0.94,0.86,0.69 and 0.82)for the training;and the AUROC of(0.76,0.81,0.85,0.72 and 0.75)for the test,subsequently.When using additional samples,the same models achieved the AUROC values of(0.92,0.94,0.88,0.75 and 0.84)for the training and(0.78,0.82,0.82,0.78 and 0.80)for the test,respectively.Using the additional samples improved the test accuracy of all the models except SVM.As a result,in data-scarce environments,this research showed that utilizing GANs to generate supplementary samples is promising because it can improve the predictive
基金supported by the National Natural Science Foundation of China(52122901,52079016)Fundamental Research Funds for the Central Universities(DUT21GJ203+1 种基金the UK Royal Society(Ref:IF160108 and IEC\NSFC\170249)sponsored by the China Scholarship Council(202106060094).
文摘Contamination events in water distribution networks(WDNs)can have a huge impact on water supply and public health;increasingly,online water quality sensors are deployed for real-time detection of contamination events.Machine learning has been used to integrate multivariate time series water quality data at multiple stations for contamination detection;however,accurate extraction of spatial features in water quality signals remains challenging.This study proposed a contamination detection method based on generative adversarial networks(GANs).The GAN model was constructed to simultaneously consider the spatial correlation between sensor locations and temporal information of water quality indicators.The model consists of two networksda generator and a discriminatordthe outputs of which are used to measure the degree of abnormality of water quality data at each time step,referred to as the anomaly score.Bayesian sequential analysis is used to update the likelihood of event occurrence based on the anomaly scores.Alarms are then generated from the fusion of single-site and multi-site models.The proposed method was tested on a WDN for various contamination events with different characteristics.Results showed high detection performance by the proposed GAN method compared with the minimum volume ellipsoid benchmark method for various contamination amplitudes.Additionally,the GAN method achieved high accuracy for various contamination events with different amplitudes and numbers of anomalous water quality parameters,and water quality data from different sensor stations,highlighting its robustness and potential for practical application to real-time contamination events.
文摘The classification of lung nodules is a challenging problem as the visual analysis of the nodules and non-nodules revealed homogenous textural patterns.In this work,an Auxiliary Classifier(AC)-Generative Adversarial Network(GAN)based Lung Cancer Classification(LCC)system is developed.The pro-posed AC-GAN-LCC system consists of three modules;preprocessing,Lungs Region Detection(LRD),and AC-GAN classification.A Wienerfilter is employed in the preprocessing module to remove the Gaussian noise.In the LRD module,only the lung regions(left and right lungs)are detected using itera-tive thresholding and morphological operations.In order to extract the lung region only,floodfilling and background subtraction.The detected lung regions are fed to the AC-GAN classifier to detect the nodules.It classifies the nodules into one of the two classes,i.e.,binary classification(such as nodules or non-nodules).The AC-GAN is the extended version of the conditional GAN that predicts the label of a given image.Three different optimization techniques,adaptive gradient optimi-zation,root mean square propagation optimization,and Adam optimization are employed for optimizing the AC-GAN architecture.The proposed AC-GAN-LCC system is evaluated on the Lung Image Database Consortium(LIDC)data-base Computed Tomography(CT)scan images.The proposed AC-GAN-LCC system classifies∼15000 CT slices(7310 non-nodules and 7685 nodules).It pro-vides an overall accuracy of 98.8%on the LIDC database using Adam optimiza-tion by a 10-fold cross-validation approach.
基金This work was supported in part by the National Natural Science Foundation of China(NSFC)(92167106,62103362,and 61833014)the Natural Science Foundation of Zhejiang Province(LR18F030001).
文摘Recently developed fault classification methods for industrial processes are mainly data-driven.Notably,models based on deep neural networks have significantly improved fault classification accuracy owing to the inclusion of a large number of data patterns.However,these data-driven models are vulnerable to adversarial attacks;thus,small perturbations on the samples can cause the models to provide incorrect fault predictions.Several recent studies have demonstrated the vulnerability of machine learning methods and the existence of adversarial samples.This paper proposes a black-box attack method with an extreme constraint for a safe-critical industrial fault classification system:Only one variable can be perturbed to craft adversarial samples.Moreover,to hide the adversarial samples in the visualization space,a Jacobian matrix is used to guide the perturbed variable selection,making the adversarial samples in the dimensional reduction space invisible to the human eye.Using the one-variable attack(OVA)method,we explore the vulnerability of industrial variables and fault types,which can help understand the geometric characteristics of fault classification systems.Based on the attack method,a corresponding adversarial training defense method is also proposed,which efficiently defends against an OVA and improves the prediction accuracy of the classifiers.In experiments,the proposed method was tested on two datasets from the Tennessee–Eastman process(TEP)and steel plates(SP).We explore the vulnerability and correlation within variables and faults and verify the effectiveness of OVAs and defenses for various classifiers and datasets.For industrial fault classification systems,the attack success rate of our method is close to(on TEP)or even higher than(on SP)the current most effective first-order white-box attack method,which requires perturbation of all variables.
