Human disturbance activities is one of the main reasons for inducing geohazards.Ecological impact assessment metrics of roads are inconsistent criteria and multiple.From the perspective of visual observation,the envir...Human disturbance activities is one of the main reasons for inducing geohazards.Ecological impact assessment metrics of roads are inconsistent criteria and multiple.From the perspective of visual observation,the environment damage can be shown through detecting the uncovered area of vegetation in the images along road.To realize this,an end-to-end environment damage detection model based on convolutional neural network is proposed.A 50-layer residual network is used to extract feature map.The initial parameters are optimized by transfer learning.An example is shown by this method.The dataset including cliff and landslide damage are collected by us along road in Shennongjia national forest park.Results show 0.4703 average precision(AP)rating for cliff damage and 0.4809 average precision(AP)rating for landslide damage.Compared with YOLOv3,our model shows a better accuracy in cliff and landslide detection although a certain amount of speed is sacrificed.展开更多
In this paper,we propose a neural network approach to learn the parameters of a class of stochastic Lotka-Volterra systems.Approximations of the mean and covariance matrix of the observational variables are obtained f...In this paper,we propose a neural network approach to learn the parameters of a class of stochastic Lotka-Volterra systems.Approximations of the mean and covariance matrix of the observational variables are obtained from the Euler-Maruyama discretization of the underlying stochastic differential equations(SDEs),based on which the loss function is built.The stochastic gradient descent method is applied in the neural network training.Numerical experiments demonstrate the effectiveness of our method.展开更多
Fault diagnosis occupies a pivotal position within the domain of machine and equipment management.Existing methods,however,often exhibit limitations in their scope of application,typically focusing on specific types o...Fault diagnosis occupies a pivotal position within the domain of machine and equipment management.Existing methods,however,often exhibit limitations in their scope of application,typically focusing on specific types of signals or faults in individual mechanical components while being constrained by data types and inherent characteristics.To address the limitations of existing methods,we propose a fault diagnosis method based on graph neural networks(GNNs)embedded with multirelationships of intrinsic mode functions(MIMF).The approach introduces a novel graph topological structure constructed from the features of intrinsic mode functions(IMFs)of monitored signals and their multirelationships.Additionally,a graph-level based fault diagnosis network model is designed to enhance feature learning capabilities for graph samples and enable flexible application across diverse signal sources and devices.Experimental validation with datasets including independent vibration signals for gear fault detection,mixed vibration signals for concurrent gear and bearing faults,and pressure signals for hydraulic cylinder leakage characterization demonstrates the model's adaptability and superior diagnostic accuracy across various types of signals and mechanical systems.展开更多
A quantum BP neural networks model with learning algorithm is proposed. First, based on the universality of single qubit rotation gate and two-qubit controlled-NOT gate, a quantum neuron model is constructed, which is...A quantum BP neural networks model with learning algorithm is proposed. First, based on the universality of single qubit rotation gate and two-qubit controlled-NOT gate, a quantum neuron model is constructed, which is composed of input, phase rotation, aggregation, reversal rotation and output. In this model, the input is described by qubits, and the output is given by the probability of the state in which （1） is observed. The phase rotation and the reversal rotation are performed by the universal quantum gates. Secondly, the quantum BP neural networks model is constructed, in which the output layer and the hide layer are quantum neurons. With the application of the gradient descent algorithm, a learning algorithm of the model is proposed, and the continuity of the model is proved. It is shown that this model and algorithm are superior to the conventional BP networks in three aspects： convergence speed, convergence rate and robustness, by two application examples of pattern recognition and function approximation.展开更多
Throught fusion of the mechanism modeling and the neural networks modeling,a compo- nent content soft-sensor,which is composed of the equilibrium calculation model for multi-component rare earth extraction and the err...Throught fusion of the mechanism modeling and the neural networks modeling,a compo- nent content soft-sensor,which is composed of the equilibrium calculation model for multi-component rare earth extraction and the error compensation model of fuzzy system,is proposed to solve the prob- lem that the component content in countercurrent rare-earth extraction process is hardly measured on-line.An industry experiment in the extraction Y process by HAB using this hybrid soft-sensor proves its effectiveness.展开更多
In this paper we study the dynamic properties and stabilities of neural networks with delay-time (which includes the time-varying case) by differential inequalities and Lyapunov function approaches. The criteria of co...In this paper we study the dynamic properties and stabilities of neural networks with delay-time (which includes the time-varying case) by differential inequalities and Lyapunov function approaches. The criteria of connective stability, robust stability, Lyapunov stability, asymptotic atability, exponential stability and Lagrange stability of neural networks with delay-time are established, and the results obtained are very useful for the design, implementation and application of adaptive learning neural networks.展开更多
Microseismic monitoring system is one of the effective methods for deep mining geo-stress monitoring.The principle of microseismic monitoring system is to analyze the mechanical parameters contained in microseismic ev...Microseismic monitoring system is one of the effective methods for deep mining geo-stress monitoring.The principle of microseismic monitoring system is to analyze the mechanical parameters contained in microseismic events for providing accurate information of rockmass.The accurate identification of microseismic events and blasts determines the timeliness and accuracy of early warning of microseismic monitoring technology.An image identification model based on Convolutional Neural Network(CNN)is established in this paper for the seismic waveforms of microseismic events and blasts.Firstly,the training set,test set,and validation set are collected,which are composed of 5250,1500,and 750 seismic waveforms of microseismic events and blasts,respectively.The classified data sets are preprocessed and input into the constructed CNN in CPU mode for training.Results show that the accuracies of microseismic events and blasts are 99.46%and 99.33%in the test set,respectively.The accuracies of microseismic events and blasts are 100%and 98.13%in the validation set,respectively.The proposed method gives superior performance when compared with existed methods.The accuracies of models using logistic regression and artificial neural network(ANN)based on the same data set are 54.43%and 67.9%in the test set,respectively.Then,the ROC curves of the three models are obtained and compared,which show that the CNN gives an absolute advantage in this classification model when the original seismic waveform are used in training the model.It not only decreases the influence of individual differences in experience,but also removes the errors induced by source and waveform parameters.It is proved that the established discriminant method improves the efficiency and accuracy of microseismic data processing for monitoring rock instability and seismicity.展开更多
An adaptive control scheme is developed for a class of single-input nonlinear systems preceded by unknown hysteresis, which is a non-differentiable and multi-value mapping nonlinearity. The controller based on the thr...An adaptive control scheme is developed for a class of single-input nonlinear systems preceded by unknown hysteresis, which is a non-differentiable and multi-value mapping nonlinearity. The controller based on the three-layer neural network （NN）, whose weights are derived from Lyapunov stability analysis, guarantees closed-loop semiglobal stability and convergence of the tracking errors to a small residual set. An example is used to confirm the effectiveness of the proposed control scheme.展开更多
Neural networks require a lot of training to understand the model of a plant or a process. Issues such as learning speed, stability, and weight convergence remain as areas of research and comparison of many training a...Neural networks require a lot of training to understand the model of a plant or a process. Issues such as learning speed, stability, and weight convergence remain as areas of research and comparison of many training algorithms. The application of neural networks to control interior permanent magnet synchronous motor using direct torque control （DTC） is discussed. A neural network is used to emulate the state selector of the DTC. The neural networks used are the back-propagation and radial basis function. To reduce the training patterns and increase the execution speed of the training process, the inputs of switching table are converted to digital signals, i.e., one bit represent the flux error, one bit the torque error, and three bits the region of stator flux. Computer simulations of the motor and neural-network system using the two approaches are presented and compared. Discussions about the back-propagation and radial basis function as the most promising training techniques are presented, giving its advantages and disadvantages. The system using back-propagation and radial basis function networks controller has quick parallel speed and high torque response.展开更多
Due to defects of time-difference of arrival localization,which influences by speed differences of various model waveforms and waveform distortion in transmitting process,a neural network technique is introduced to ca...Due to defects of time-difference of arrival localization,which influences by speed differences of various model waveforms and waveform distortion in transmitting process,a neural network technique is introduced to calculate localization of the acoustic emission source.However,in back propagation（BP） neural network,the BP algorithm is a stochastic gradient algorithm virtually,the network may get into local minimum and the result of network training is dissatisfactory.It is a kind of genetic algorithms with the form of quantum chromosomes,the random observation which simulates the quantum collapse can bring diverse individuals,and the evolutionary operators characterized by a quantum mechanism are introduced to speed up convergence and avoid prematurity.Simulation results show that the modeling of neural network based on quantum genetic algorithm has fast convergent and higher localization accuracy,so it has a good application prospect and is worth researching further more.展开更多
Background Coronary artery calcification is a well-known marker of atherosclerotic plaque burden.High-resolution intravascular optical coherence tomography(OCT)imaging has shown the potential to characterize the detai...Background Coronary artery calcification is a well-known marker of atherosclerotic plaque burden.High-resolution intravascular optical coherence tomography(OCT)imaging has shown the potential to characterize the details of coronary calcification in vivo.In routine clinical practice,it is a time-consuming and laborious task for clinicians to review the over 250 images in a single pullback.Besides,the imbalance label distribution within the entire pullbacks is another problem,which could lead to the failure of the classifier model.Given the success of deep learning methods with other imaging modalities,a thorough understanding of calcified plaque detection using Convolutional Neural Networks(CNNs)within pullbacks for future clinical decision was required.Methods All 33 IVOCT clinical pullbacks of 33 patients were taken from Affiliated Drum Tower Hospital,Nanjing University between December 2017 and December 2018.For ground-truth annotation,three trained experts determined the type of plaque that was present in a B-Scan.The experts assigned the labels'no calcified plaque','calcified plaque'for each OCT image.All experts were provided the all images for labeling.The final label was determined based on consensus between the experts,different opinions on the plaque type were resolved by asking the experts for a repetition of their evaluation.Before the implement of algorithm,all OCT images was resized to a resolution of 300×300,which matched the range used with standard architectures in the natural image domain.In the study,we randomly selected 26 pullbacks for training,the remaining data were testing.While,imbalance label distribution within entire pullbacks was great challenge for various CNNs architecture.In order to resolve the problem,we designed the following experiment.First,we fine-tuned twenty different CNNs architecture,including customize CNN architectures and pretrained CNN architectures.Considering the nature of OCT images,customize CNN architectures were designed that the layers were fewer than 25 layers.Then,three with good performance were selected and further deep fine-tuned to train three different models.The difference of CNNs was mainly in the model architecture,such as depth-based residual networks,width-based inception networks.Finally,the three CNN models were used to majority voting,the predicted labels were from the most voting.Areas under the receiver operating characteristic curve(ROC AUC)were used as the evaluation metric for the imbalance label distribution.Results The imbalance label distribution within pullbacks affected both convergence during the training phase and generalization of a CNN model.Different labels of OCT images could be classified with excellent performance by fine tuning parameters of CNN architectures.Overall,we find that our final result performed best with an accuracy of 90%of'calcified plaque'class,which the numbers were less than'no calcified plaque'class in one pullback.Conclusions The obtained results showed that the method is fast and effective to classify calcific plaques with imbalance label distribution in each pullback.The results suggest that the proposed method could be facilitating our understanding of coronary artery calcification in the process of atherosclerosis andhelping guide complex interventional strategies in coronary arteries with superficial calcification.展开更多
The problem of passivity analysis is investigated for uncertain stochastic neural networks with discrete interval and distributed time-varying delays.The parameter uncertainties are assumed to be norm bounded and the ...The problem of passivity analysis is investigated for uncertain stochastic neural networks with discrete interval and distributed time-varying delays.The parameter uncertainties are assumed to be norm bounded and the delay is assumed to be time-varying and belongs to a given interval,which means that the lower and upper bounds of interval time-varying delays are available.By constructing proper Lyapunov-Krasovskii functional and employing a combination of the free-weighting matrix method and stochastic analysis technique,new delay-dependent passivity conditions are derived in terms of linear matrix inequalities（LMIs）.Finally,numerical examples are given to show the less conservatism of the proposed conditions.展开更多
Learning is one of key problems of artificial neural networks. In this paper, we present a kind of combined learning algorithm based on fuzzy entropy criterion for neural networks. The basic idea is to simulate the le...Learning is one of key problems of artificial neural networks. In this paper, we present a kind of combined learning algorithm based on fuzzy entropy criterion for neural networks. The basic idea is to simulate the learning mechanism of human brain and overcome the limitations of monocrifsterion learning. The comparison is made between the given learning algorithm and the typical BP algorithm in order to show the characteristics of the new algorithm.展开更多
The performance of the model algorithm control method is partially based on the accuracy of the system＇s model. It is difficult to obtain a good model of a nonlinear system, especially when the nonlinearity is high. ...The performance of the model algorithm control method is partially based on the accuracy of the system＇s model. It is difficult to obtain a good model of a nonlinear system, especially when the nonlinearity is high. Neural networks have the ability to ＂learn＂the characteristics of a system through nonlinear mapping to represent nonlinear functions as well as their inverse functions. This paper presents a model algorithm control method using neural networks for nonlinear time delay systems. Two neural networks are used in the control scheme. One neural network is trained as the model of the nonlinear time delay system, and the other one produces the control inputs. The neural networks are combined with the model algorithm control method to control the nonlinear time delay systems. Three examples are used to illustrate the proposed control method. The simulation results show that the proposed control method has a good control performance for nonlinear time delay systems.展开更多
In practical multi-sensor information fusion systems, there exists uncertainty about the network structure, active state of sensors, and information itself （including fuzziness, randomness, incompleteness as well as ...In practical multi-sensor information fusion systems, there exists uncertainty about the network structure, active state of sensors, and information itself （including fuzziness, randomness, incompleteness as well as roughness, etc）. Hence it requires investigating the problem of uncertain information fusion. Robust learning algorithm which adapts to complex environment and the fuzzy inference algorithm which disposes fuzzy information are explored to solve the problem. Based on the fusion technology of neural networks and fuzzy inference algorithm, a multi-sensor uncertain information fusion system is modeled. Also RANFIS learning algorithm and fusing weight synthesized inference algorithm are developed from the ANFIS algorithm according to the concept of robust neural networks. This fusion system mainly consists of RANFIS confidence estimator, fusing weight synthesized inference knowledge base and weighted fusion section. The simulation result demonstrates that the proposed fusion model and algorithm have the capability of uncertain information fusion, thus is obviously advantageous compared with the conventional Kalman weighted fusion algorithm.展开更多
This paper presents an automatic system of vehicle number-plate recognition based on neural networks. In this system, location of number-plate and recognition of characters in number-plate can be automatically complet...This paper presents an automatic system of vehicle number-plate recognition based on neural networks. In this system, location of number-plate and recognition of characters in number-plate can be automatically completed. Pixel colors of Number-plate area are classified using neural network, then color features are extracted by analyzing scanning lines of the cross-section of number-plate. It takes full use of number-plate color features to locate number-plate. Characters in number-plate can be effectively recognized using the neural networks. Experimental results show that the correct rate of number-plate location is close to 100%, and the time of number-plate location is less than 1 second. Moreover, recognition rate of characters is improved due to the known number-plate type. It is also observed that this system is not sensitive to variations of weather, illumination and vehicle speed. In addition, and also the size of number-plate need not to be known in prior. This system is of crucial significance to apply and spread the automatic system of vehicle number-plate recognition.展开更多
Newton＇s learning algorithm of NN is presented and realized. In theory, the convergence rate of learning algorithm of NN based on Newton＇s method must be faster than BP＇s and other learning algorithms, because the ...Newton＇s learning algorithm of NN is presented and realized. In theory, the convergence rate of learning algorithm of NN based on Newton＇s method must be faster than BP＇s and other learning algorithms, because the gradient method is linearly convergent while Newton＇s method has second order convergence rate. The fast computing algorithm of Hesse matrix of the cost function of NN is proposed and it is the theory basis of the improvement of Newton＇s learning algorithm. Simulation results show that the convergence rate of Newton＇s learning algorithm is high and apparently faster than the traditional BP method＇s, and the robustness of Newton＇s learning algorithm is also better than BP method＇ s.展开更多
The Volterra feedforward neural network with nonlinear interconnections and related homotopy learning algorithm are proposed in the paper. It is shown that Volterra neural network and the homolopy learning algorithms ...The Volterra feedforward neural network with nonlinear interconnections and related homotopy learning algorithm are proposed in the paper. It is shown that Volterra neural network and the homolopy learning algorithms are significant potentials in nonlinear approximation ability,convergent speeds and global optimization than the classical neural networks and the standard BP algorithm, and related computer simulations and theoretical analysis are given too.展开更多
The Radial Basis Functions Neural Network （RBFNN） is used to establish the model of a response system through the input and output data of the system. The synchronization between a drive system and the response syst...The Radial Basis Functions Neural Network （RBFNN） is used to establish the model of a response system through the input and output data of the system. The synchronization between a drive system and the response system can be implemented by employing the RBFNN model and state feedback control. In this case, the exact mathematical model, which is the precondition for the conventional method, is unnecessary for implementing synchronization. The effect of the model error is investigated and a corresponding theorem is developed. The effect of the parameter perturbations and the measurement noise is investigated through simulations. The simulation results under different conditions show the effectiveness of the method.展开更多
文摘Human disturbance activities is one of the main reasons for inducing geohazards.Ecological impact assessment metrics of roads are inconsistent criteria and multiple.From the perspective of visual observation,the environment damage can be shown through detecting the uncovered area of vegetation in the images along road.To realize this,an end-to-end environment damage detection model based on convolutional neural network is proposed.A 50-layer residual network is used to extract feature map.The initial parameters are optimized by transfer learning.An example is shown by this method.The dataset including cliff and landslide damage are collected by us along road in Shennongjia national forest park.Results show 0.4703 average precision(AP)rating for cliff damage and 0.4809 average precision(AP)rating for landslide damage.Compared with YOLOv3,our model shows a better accuracy in cliff and landslide detection although a certain amount of speed is sacrificed.
基金Supported by the National Natural Science Foundation of China(11971458,11471310)。
文摘In this paper,we propose a neural network approach to learn the parameters of a class of stochastic Lotka-Volterra systems.Approximations of the mean and covariance matrix of the observational variables are obtained from the Euler-Maruyama discretization of the underlying stochastic differential equations(SDEs),based on which the loss function is built.The stochastic gradient descent method is applied in the neural network training.Numerical experiments demonstrate the effectiveness of our method.
文摘Fault diagnosis occupies a pivotal position within the domain of machine and equipment management.Existing methods,however,often exhibit limitations in their scope of application,typically focusing on specific types of signals or faults in individual mechanical components while being constrained by data types and inherent characteristics.To address the limitations of existing methods,we propose a fault diagnosis method based on graph neural networks(GNNs)embedded with multirelationships of intrinsic mode functions(MIMF).The approach introduces a novel graph topological structure constructed from the features of intrinsic mode functions(IMFs)of monitored signals and their multirelationships.Additionally,a graph-level based fault diagnosis network model is designed to enhance feature learning capabilities for graph samples and enable flexible application across diverse signal sources and devices.Experimental validation with datasets including independent vibration signals for gear fault detection,mixed vibration signals for concurrent gear and bearing faults,and pressure signals for hydraulic cylinder leakage characterization demonstrates the model's adaptability and superior diagnostic accuracy across various types of signals and mechanical systems.
基金the National Natural Science Foundation of China (50138010)
文摘A quantum BP neural networks model with learning algorithm is proposed. First, based on the universality of single qubit rotation gate and two-qubit controlled-NOT gate, a quantum neuron model is constructed, which is composed of input, phase rotation, aggregation, reversal rotation and output. In this model, the input is described by qubits, and the output is given by the probability of the state in which （1） is observed. The phase rotation and the reversal rotation are performed by the universal quantum gates. Secondly, the quantum BP neural networks model is constructed, in which the output layer and the hide layer are quantum neurons. With the application of the gradient descent algorithm, a learning algorithm of the model is proposed, and the continuity of the model is proved. It is shown that this model and algorithm are superior to the conventional BP networks in three aspects： convergence speed, convergence rate and robustness, by two application examples of pattern recognition and function approximation.
基金Supported by National Natural Science Foundation of P.R.China(50474020,60534010,60504006)
文摘Throught fusion of the mechanism modeling and the neural networks modeling,a compo- nent content soft-sensor,which is composed of the equilibrium calculation model for multi-component rare earth extraction and the error compensation model of fuzzy system,is proposed to solve the prob- lem that the component content in countercurrent rare-earth extraction process is hardly measured on-line.An industry experiment in the extraction Y process by HAB using this hybrid soft-sensor proves its effectiveness.
文摘In this paper we study the dynamic properties and stabilities of neural networks with delay-time (which includes the time-varying case) by differential inequalities and Lyapunov function approaches. The criteria of connective stability, robust stability, Lyapunov stability, asymptotic atability, exponential stability and Lagrange stability of neural networks with delay-time are established, and the results obtained are very useful for the design, implementation and application of adaptive learning neural networks.
基金Projects(51822407,51774327,51664016)supported by the National Natural Science Foundation of China。
文摘Microseismic monitoring system is one of the effective methods for deep mining geo-stress monitoring.The principle of microseismic monitoring system is to analyze the mechanical parameters contained in microseismic events for providing accurate information of rockmass.The accurate identification of microseismic events and blasts determines the timeliness and accuracy of early warning of microseismic monitoring technology.An image identification model based on Convolutional Neural Network(CNN)is established in this paper for the seismic waveforms of microseismic events and blasts.Firstly,the training set,test set,and validation set are collected,which are composed of 5250,1500,and 750 seismic waveforms of microseismic events and blasts,respectively.The classified data sets are preprocessed and input into the constructed CNN in CPU mode for training.Results show that the accuracies of microseismic events and blasts are 99.46%and 99.33%in the test set,respectively.The accuracies of microseismic events and blasts are 100%and 98.13%in the validation set,respectively.The proposed method gives superior performance when compared with existed methods.The accuracies of models using logistic regression and artificial neural network(ANN)based on the same data set are 54.43%and 67.9%in the test set,respectively.Then,the ROC curves of the three models are obtained and compared,which show that the CNN gives an absolute advantage in this classification model when the original seismic waveform are used in training the model.It not only decreases the influence of individual differences in experience,but also removes the errors induced by source and waveform parameters.It is proved that the established discriminant method improves the efficiency and accuracy of microseismic data processing for monitoring rock instability and seismicity.
文摘An adaptive control scheme is developed for a class of single-input nonlinear systems preceded by unknown hysteresis, which is a non-differentiable and multi-value mapping nonlinearity. The controller based on the three-layer neural network （NN）, whose weights are derived from Lyapunov stability analysis, guarantees closed-loop semiglobal stability and convergence of the tracking errors to a small residual set. An example is used to confirm the effectiveness of the proposed control scheme.
基金the National Natural Science Foundation of China (60374032).
文摘Neural networks require a lot of training to understand the model of a plant or a process. Issues such as learning speed, stability, and weight convergence remain as areas of research and comparison of many training algorithms. The application of neural networks to control interior permanent magnet synchronous motor using direct torque control （DTC） is discussed. A neural network is used to emulate the state selector of the DTC. The neural networks used are the back-propagation and radial basis function. To reduce the training patterns and increase the execution speed of the training process, the inputs of switching table are converted to digital signals, i.e., one bit represent the flux error, one bit the torque error, and three bits the region of stator flux. Computer simulations of the motor and neural-network system using the two approaches are presented and compared. Discussions about the back-propagation and radial basis function as the most promising training techniques are presented, giving its advantages and disadvantages. The system using back-propagation and radial basis function networks controller has quick parallel speed and high torque response.
基金supported by the National Natural Science Foundation of China (51075068)the Southeast University Science Foundation Funded Program (KJ2009348)
文摘Due to defects of time-difference of arrival localization,which influences by speed differences of various model waveforms and waveform distortion in transmitting process,a neural network technique is introduced to calculate localization of the acoustic emission source.However,in back propagation（BP） neural network,the BP algorithm is a stochastic gradient algorithm virtually,the network may get into local minimum and the result of network training is dissatisfactory.It is a kind of genetic algorithms with the form of quantum chromosomes,the random observation which simulates the quantum collapse can bring diverse individuals,and the evolutionary operators characterized by a quantum mechanism are introduced to speed up convergence and avoid prematurity.Simulation results show that the modeling of neural network based on quantum genetic algorithm has fast convergent and higher localization accuracy,so it has a good application prospect and is worth researching further more.
基金supported in part by the National Natural Science Foundation of China ( NSFC ) ( 11772093)ARC ( FT140101152)
文摘Background Coronary artery calcification is a well-known marker of atherosclerotic plaque burden.High-resolution intravascular optical coherence tomography(OCT)imaging has shown the potential to characterize the details of coronary calcification in vivo.In routine clinical practice,it is a time-consuming and laborious task for clinicians to review the over 250 images in a single pullback.Besides,the imbalance label distribution within the entire pullbacks is another problem,which could lead to the failure of the classifier model.Given the success of deep learning methods with other imaging modalities,a thorough understanding of calcified plaque detection using Convolutional Neural Networks(CNNs)within pullbacks for future clinical decision was required.Methods All 33 IVOCT clinical pullbacks of 33 patients were taken from Affiliated Drum Tower Hospital,Nanjing University between December 2017 and December 2018.For ground-truth annotation,three trained experts determined the type of plaque that was present in a B-Scan.The experts assigned the labels'no calcified plaque','calcified plaque'for each OCT image.All experts were provided the all images for labeling.The final label was determined based on consensus between the experts,different opinions on the plaque type were resolved by asking the experts for a repetition of their evaluation.Before the implement of algorithm,all OCT images was resized to a resolution of 300×300,which matched the range used with standard architectures in the natural image domain.In the study,we randomly selected 26 pullbacks for training,the remaining data were testing.While,imbalance label distribution within entire pullbacks was great challenge for various CNNs architecture.In order to resolve the problem,we designed the following experiment.First,we fine-tuned twenty different CNNs architecture,including customize CNN architectures and pretrained CNN architectures.Considering the nature of OCT images,customize CNN architectures were designed that the layers were fewer than 25 layers.Then,three with good performance were selected and further deep fine-tuned to train three different models.The difference of CNNs was mainly in the model architecture,such as depth-based residual networks,width-based inception networks.Finally,the three CNN models were used to majority voting,the predicted labels were from the most voting.Areas under the receiver operating characteristic curve(ROC AUC)were used as the evaluation metric for the imbalance label distribution.Results The imbalance label distribution within pullbacks affected both convergence during the training phase and generalization of a CNN model.Different labels of OCT images could be classified with excellent performance by fine tuning parameters of CNN architectures.Overall,we find that our final result performed best with an accuracy of 90%of'calcified plaque'class,which the numbers were less than'no calcified plaque'class in one pullback.Conclusions The obtained results showed that the method is fast and effective to classify calcific plaques with imbalance label distribution in each pullback.The results suggest that the proposed method could be facilitating our understanding of coronary artery calcification in the process of atherosclerosis andhelping guide complex interventional strategies in coronary arteries with superficial calcification.
基金supported by Department of Science and Technology,New Delhi,India(SR/S4/MS:485/07)
文摘The problem of passivity analysis is investigated for uncertain stochastic neural networks with discrete interval and distributed time-varying delays.The parameter uncertainties are assumed to be norm bounded and the delay is assumed to be time-varying and belongs to a given interval,which means that the lower and upper bounds of interval time-varying delays are available.By constructing proper Lyapunov-Krasovskii functional and employing a combination of the free-weighting matrix method and stochastic analysis technique,new delay-dependent passivity conditions are derived in terms of linear matrix inequalities（LMIs）.Finally,numerical examples are given to show the less conservatism of the proposed conditions.
文摘Learning is one of key problems of artificial neural networks. In this paper, we present a kind of combined learning algorithm based on fuzzy entropy criterion for neural networks. The basic idea is to simulate the learning mechanism of human brain and overcome the limitations of monocrifsterion learning. The comparison is made between the given learning algorithm and the typical BP algorithm in order to show the characteristics of the new algorithm.
基金supported by the Brain Korea 21 PLUS Project,National Research Foundation of Korea(NRF-2013R1A2A2A01068127NRF-2013R1A1A2A10009458)Jiangsu Province University Natural Science Research Project(13KJB510003)
文摘The performance of the model algorithm control method is partially based on the accuracy of the system＇s model. It is difficult to obtain a good model of a nonlinear system, especially when the nonlinearity is high. Neural networks have the ability to ＂learn＂the characteristics of a system through nonlinear mapping to represent nonlinear functions as well as their inverse functions. This paper presents a model algorithm control method using neural networks for nonlinear time delay systems. Two neural networks are used in the control scheme. One neural network is trained as the model of the nonlinear time delay system, and the other one produces the control inputs. The neural networks are combined with the model algorithm control method to control the nonlinear time delay systems. Three examples are used to illustrate the proposed control method. The simulation results show that the proposed control method has a good control performance for nonlinear time delay systems.
基金This project was supported by the National Natural Science Foundation of China (60572038)
文摘In practical multi-sensor information fusion systems, there exists uncertainty about the network structure, active state of sensors, and information itself （including fuzziness, randomness, incompleteness as well as roughness, etc）. Hence it requires investigating the problem of uncertain information fusion. Robust learning algorithm which adapts to complex environment and the fuzzy inference algorithm which disposes fuzzy information are explored to solve the problem. Based on the fusion technology of neural networks and fuzzy inference algorithm, a multi-sensor uncertain information fusion system is modeled. Also RANFIS learning algorithm and fusing weight synthesized inference algorithm are developed from the ANFIS algorithm according to the concept of robust neural networks. This fusion system mainly consists of RANFIS confidence estimator, fusing weight synthesized inference knowledge base and weighted fusion section. The simulation result demonstrates that the proposed fusion model and algorithm have the capability of uncertain information fusion, thus is obviously advantageous compared with the conventional Kalman weighted fusion algorithm.
基金the Key Development Plan Project of Science and Technology (No.991p0111).
文摘This paper presents an automatic system of vehicle number-plate recognition based on neural networks. In this system, location of number-plate and recognition of characters in number-plate can be automatically completed. Pixel colors of Number-plate area are classified using neural network, then color features are extracted by analyzing scanning lines of the cross-section of number-plate. It takes full use of number-plate color features to locate number-plate. Characters in number-plate can be effectively recognized using the neural networks. Experimental results show that the correct rate of number-plate location is close to 100%, and the time of number-plate location is less than 1 second. Moreover, recognition rate of characters is improved due to the known number-plate type. It is also observed that this system is not sensitive to variations of weather, illumination and vehicle speed. In addition, and also the size of number-plate need not to be known in prior. This system is of crucial significance to apply and spread the automatic system of vehicle number-plate recognition.
文摘Newton＇s learning algorithm of NN is presented and realized. In theory, the convergence rate of learning algorithm of NN based on Newton＇s method must be faster than BP＇s and other learning algorithms, because the gradient method is linearly convergent while Newton＇s method has second order convergence rate. The fast computing algorithm of Hesse matrix of the cost function of NN is proposed and it is the theory basis of the improvement of Newton＇s learning algorithm. Simulation results show that the convergence rate of Newton＇s learning algorithm is high and apparently faster than the traditional BP method＇s, and the robustness of Newton＇s learning algorithm is also better than BP method＇ s.
文摘The Volterra feedforward neural network with nonlinear interconnections and related homotopy learning algorithm are proposed in the paper. It is shown that Volterra neural network and the homolopy learning algorithms are significant potentials in nonlinear approximation ability,convergent speeds and global optimization than the classical neural networks and the standard BP algorithm, and related computer simulations and theoretical analysis are given too.
基金This project was supported in part by the Science Foundation of Shanxi Province (2003F028)China Postdoctoral Science Foundation (20060390318).
文摘The Radial Basis Functions Neural Network （RBFNN） is used to establish the model of a response system through the input and output data of the system. The synchronization between a drive system and the response system can be implemented by employing the RBFNN model and state feedback control. In this case, the exact mathematical model, which is the precondition for the conventional method, is unnecessary for implementing synchronization. The effect of the model error is investigated and a corresponding theorem is developed. The effect of the parameter perturbations and the measurement noise is investigated through simulations. The simulation results under different conditions show the effectiveness of the method.
