The relationship among Mercer kernel, reproducing kernel and positive definite kernel in support vector machine （SVM） is proved and their roles in SVM are discussed. The quadratic form of the kernel matrix is used t...The relationship among Mercer kernel, reproducing kernel and positive definite kernel in support vector machine （SVM） is proved and their roles in SVM are discussed. The quadratic form of the kernel matrix is used to confirm the positive definiteness and their construction. Based on the Bochner theorem, some translation invariant kernels are checked in their Fourier domain. Some rotation invariant radial kernels are inspected according to the Schoenberg theorem. Finally, the construction of discrete scaling and wavelet kernels, the kernel selection and the kernel parameter learning are discussed.展开更多
The distribution of data has a significant impact on the results of classification.When the distribution of one class is insignificant compared to the distribution of another class,data imbalance occurs.This will resu...The distribution of data has a significant impact on the results of classification.When the distribution of one class is insignificant compared to the distribution of another class,data imbalance occurs.This will result in rising outlier values and noise.Therefore,the speed and performance of classification could be greatly affected.Given the above problems,this paper starts with the motivation and mathematical representing of classification,puts forward a new classification method based on the relationship between different classification formulations.Combined with the vector characteristics of the actual problem and the choice of matrix characteristics,we firstly analyze the orderly regression to introduce slack variables to solve the constraint problem of the lone point.Then we introduce the fuzzy factors to solve the problem of the gap between the isolated points on the basis of the support vector machine.We introduce the cost control to solve the problem of sample skew.Finally,based on the bi-boundary support vector machine,a twostep weight setting twin classifier is constructed.This can help to identify multitasks with feature-selected patterns without the need for additional optimizers,which solves the problem of large-scale classification that can’t deal effectively with the very low category distribution gap.展开更多
Support vector machines （SVM） have been widely used in chaotic time series predictions in recent years. In order to enhance the prediction efficiency of this method and implement it in hardware, the sigmoid kernel i...Support vector machines （SVM） have been widely used in chaotic time series predictions in recent years. In order to enhance the prediction efficiency of this method and implement it in hardware, the sigmoid kernel in SVM is drawn in a more natural way by using the fuzzy logic method proposed in this paper. This method provides easy hardware implementation and straightforward interpretability. Experiments on two typical chaotic time series predictions have been carried out and the obtained results show that the average CPU time can be reduced significantly at the cost of a small decrease in prediction accuracy, which is favourable for the hardware implementation for chaotic time series prediction.展开更多
Support vector machine(SVM)is easily affected by noises and outliers,and its training time dramatically increases with the growing in number of training samples.Satellite cloud image may easily be deteriorated by nois...Support vector machine(SVM)is easily affected by noises and outliers,and its training time dramatically increases with the growing in number of training samples.Satellite cloud image may easily be deteriorated by noises and intensity non-uniformity with a huge amount of data needs to be processed regularly,so it is hard to detect convective clouds in satellite image using traditional SVM.To deal with this problem,a novel method for detection of convective clouds was proposed based on fast fuzzy support vector machine(FFSVM).FFSVM was constructed by eliminating feeble samples and designing new membership function as two aspects.Firstly,according to the distribution characteristics of fuzzy inseparable sample set and the fact that the classification hyper-plane is only determined by support vectors,this paper uses SVDD,Gaussian model and border vector extraction model comprehensively to design a sample selection method in three steps,which can eliminate most of redundant samples and keep possible support vectors.Then,by defining adaptive parameters related to attenuation rate and critical membership on the basis of the distribution characteristics of training set,an adaptive membership function is designed.Finally,the FFSVM is trained by the remaining samples using adaptive membership function to detect convective clouds.The experiments on FY-2D satellite images show that the proposed method,compared with traditional FSVM,not only remarkably reduces training time,but also further improves the accuracy of convective clouds detection.展开更多
The support vector machine(SVM)is a classical machine learning method.Both the hinge loss and least absolute shrinkage and selection operator(LASSO)penalty are usually used in traditional SVMs.However,the hinge loss i...The support vector machine(SVM)is a classical machine learning method.Both the hinge loss and least absolute shrinkage and selection operator(LASSO)penalty are usually used in traditional SVMs.However,the hinge loss is not differentiable,and the LASSO penalty does not have the Oracle property.In this paper,the huberized loss is combined with non-convex penalties to obtain a model that has the advantages of both the computational simplicity and the Oracle property,contributing to higher accuracy than traditional SVMs.It is experimentally demonstrated that the two non-convex huberized-SVM methods,smoothly clipped absolute deviation huberized-SVM(SCAD-HSVM)and minimax concave penalty huberized-SVM(MCP-HSVM),outperform the traditional SVM method in terms of the prediction accuracy and classifier performance.They are also superior in terms of variable selection,especially when there is a high linear correlation between the variables.When they are applied to the prediction of listed companies,the variables that can affect and predict financial distress are accurately filtered out.Among all the indicators,the indicators per share have the greatest influence while those of solvency have the weakest influence.Listed companies can assess the financial situation with the indicators screened by our algorithm and make an early warning of their possible financial distress in advance with higher precision.展开更多
Laser-induced breakdown spectroscopy（LIBS） is a versatile tool for both qualitative and quantitative analysis.In this paper,LIBS combined with principal component analysis（PCA） and support vector machine（SVM） is...Laser-induced breakdown spectroscopy（LIBS） is a versatile tool for both qualitative and quantitative analysis.In this paper,LIBS combined with principal component analysis（PCA） and support vector machine（SVM） is applied to rock analysis.Fourteen emission lines including Fe,Mg,Ca,Al,Si,and Ti are selected as analysis lines.A good accuracy（91.38% for the real rock） is achieved by using SVM to analyze the spectroscopic peak area data which are processed by PCA.It can not only reduce the noise and dimensionality which contributes to improving the efficiency of the program,but also solve the problem of linear inseparability by combining PCA and SVM.By this method,the ability of LIBS to classify rock is validated.展开更多
A method of applying support vector machine (SVM) in speech recognition was proposed, and a speech recognition system for mandarin digits was built up by SVMs. In the system, vectors were linearly extracted from speec...A method of applying support vector machine (SVM) in speech recognition was proposed, and a speech recognition system for mandarin digits was built up by SVMs. In the system, vectors were linearly extracted from speech feature sequence to make up time-aligned input patterns for SVM, and the decisions of several 2-class SVM classifiers were employed for constructing an N-class classifier. Four kinds of SVM kernel functions were compared in the experiments of speaker-independent speech recognition of mandarin digits. And the kernel of radial basis function has the highest accurate rate of 99.33%, which is better than that of the baseline system based on hidden Markov models (HMM) (97.08%). And the experiments also show that SVM can outperform HMM especially when the samples for learning were very limited.展开更多
An adaptive blind support vector machine equalizer（ABSVME） is presented in this paper.The method is based upon least square support vector machine（LSSVM）,and stems from signal feature reconstruction idea.By oversa...An adaptive blind support vector machine equalizer（ABSVME） is presented in this paper.The method is based upon least square support vector machine（LSSVM）,and stems from signal feature reconstruction idea.By oversampling the output of a LSSVM equalizer and exploiting a reasonable decorrelation cost function design,the method achieves fine online channel tracing with Kumar express algorithm and static iterative learning algorithm incorporated.The method is verified through simulation and compared with other nonlinear equalizers.The results show that it provides excellent performance in nonlinear equalization and time-varying channel tracing.Although a constant module equalization algorithm requires that the signal has characteristic of constant module,this method has no such requirement.展开更多
In order to solve the problems of small sample over-fitting and local minima when neural networks learn online, a novel method of predicting network bandwidth based on support vector machines(SVM) is proposed. The pre...In order to solve the problems of small sample over-fitting and local minima when neural networks learn online, a novel method of predicting network bandwidth based on support vector machines(SVM) is proposed. The prediction and learning online will be completed by the proposed moving window learning algorithm(MWLA). The simulation research is done to validate the proposed method, which is compared with the method based on neural networks.展开更多
To improve the training speed of support vector machine （SVM）, a method called improved center distance ratio method （ICDRM） with determining thresholds automatically is presented here without reduce the identific...To improve the training speed of support vector machine （SVM）, a method called improved center distance ratio method （ICDRM） with determining thresholds automatically is presented here without reduce the identification rate. In this method border vectors are chosen from the given samples by comparing sample vectors with center distance ratio in advance. The number of training samples is reduced greatly and the training speed is improved. This method is used to the identification for license plate characters. Experimental resuhs show that the improved SVM method-ICDRM does well at identification rate and training speed.展开更多
基金Supported by the National Natural Science Foundation of China(60473035)~~
文摘The relationship among Mercer kernel, reproducing kernel and positive definite kernel in support vector machine （SVM） is proved and their roles in SVM are discussed. The quadratic form of the kernel matrix is used to confirm the positive definiteness and their construction. Based on the Bochner theorem, some translation invariant kernels are checked in their Fourier domain. Some rotation invariant radial kernels are inspected according to the Schoenberg theorem. Finally, the construction of discrete scaling and wavelet kernels, the kernel selection and the kernel parameter learning are discussed.
基金Hebei Province Key Research and Development Project(No.20313701D)Hebei Province Key Research and Development Project(No.19210404D)+13 种基金Mobile computing and universal equipment for the Beijing Key Laboratory Open Project,The National Social Science Fund of China(17AJL014)Beijing University of Posts and Telecommunications Construction of World-Class Disciplines and Characteristic Development Guidance Special Fund “Cultural Inheritance and Innovation”Project(No.505019221)National Natural Science Foundation of China(No.U1536112)National Natural Science Foundation of China(No.81673697)National Natural Science Foundation of China(61872046)The National Social Science Fund Key Project of China(No.17AJL014)“Blue Fire Project”(Huizhou)University of Technology Joint Innovation Project(CXZJHZ201729)Industry-University Cooperation Cooperative Education Project of the Ministry of Education(No.201902218004)Industry-University Cooperation Cooperative Education Project of the Ministry of Education(No.201902024006)Industry-University Cooperation Cooperative Education Project of the Ministry of Education(No.201901197007)Industry-University Cooperation Collaborative Education Project of the Ministry of Education(No.201901199005)The Ministry of Education Industry-University Cooperation Collaborative Education Project(No.201901197001)Shijiazhuang science and technology plan project(236240267A)Hebei Province key research and development plan project(20312701D)。
文摘The distribution of data has a significant impact on the results of classification.When the distribution of one class is insignificant compared to the distribution of another class,data imbalance occurs.This will result in rising outlier values and noise.Therefore,the speed and performance of classification could be greatly affected.Given the above problems,this paper starts with the motivation and mathematical representing of classification,puts forward a new classification method based on the relationship between different classification formulations.Combined with the vector characteristics of the actual problem and the choice of matrix characteristics,we firstly analyze the orderly regression to introduce slack variables to solve the constraint problem of the lone point.Then we introduce the fuzzy factors to solve the problem of the gap between the isolated points on the basis of the support vector machine.We introduce the cost control to solve the problem of sample skew.Finally,based on the bi-boundary support vector machine,a twostep weight setting twin classifier is constructed.This can help to identify multitasks with feature-selected patterns without the need for additional optimizers,which solves the problem of large-scale classification that can’t deal effectively with the very low category distribution gap.
文摘Support vector machines （SVM） have been widely used in chaotic time series predictions in recent years. In order to enhance the prediction efficiency of this method and implement it in hardware, the sigmoid kernel in SVM is drawn in a more natural way by using the fuzzy logic method proposed in this paper. This method provides easy hardware implementation and straightforward interpretability. Experiments on two typical chaotic time series predictions have been carried out and the obtained results show that the average CPU time can be reduced significantly at the cost of a small decrease in prediction accuracy, which is favourable for the hardware implementation for chaotic time series prediction.
基金supported in part by the National Natural Science Foundation of China under Grants (61471212)Natural Science Foundation of Zhejiang Province under Grants (LY16F010001)+1 种基金Science and Technology Program of Zhejiang Meteorological Bureau under Grants (2016YB01)Natural Science Foundation of Ningbo under Grants(2016A610091,2017A610297)
文摘Support vector machine(SVM)is easily affected by noises and outliers,and its training time dramatically increases with the growing in number of training samples.Satellite cloud image may easily be deteriorated by noises and intensity non-uniformity with a huge amount of data needs to be processed regularly,so it is hard to detect convective clouds in satellite image using traditional SVM.To deal with this problem,a novel method for detection of convective clouds was proposed based on fast fuzzy support vector machine(FFSVM).FFSVM was constructed by eliminating feeble samples and designing new membership function as two aspects.Firstly,according to the distribution characteristics of fuzzy inseparable sample set and the fact that the classification hyper-plane is only determined by support vectors,this paper uses SVDD,Gaussian model and border vector extraction model comprehensively to design a sample selection method in three steps,which can eliminate most of redundant samples and keep possible support vectors.Then,by defining adaptive parameters related to attenuation rate and critical membership on the basis of the distribution characteristics of training set,an adaptive membership function is designed.Finally,the FFSVM is trained by the remaining samples using adaptive membership function to detect convective clouds.The experiments on FY-2D satellite images show that the proposed method,compared with traditional FSVM,not only remarkably reduces training time,but also further improves the accuracy of convective clouds detection.
文摘The support vector machine(SVM)is a classical machine learning method.Both the hinge loss and least absolute shrinkage and selection operator(LASSO)penalty are usually used in traditional SVMs.However,the hinge loss is not differentiable,and the LASSO penalty does not have the Oracle property.In this paper,the huberized loss is combined with non-convex penalties to obtain a model that has the advantages of both the computational simplicity and the Oracle property,contributing to higher accuracy than traditional SVMs.It is experimentally demonstrated that the two non-convex huberized-SVM methods,smoothly clipped absolute deviation huberized-SVM(SCAD-HSVM)and minimax concave penalty huberized-SVM(MCP-HSVM),outperform the traditional SVM method in terms of the prediction accuracy and classifier performance.They are also superior in terms of variable selection,especially when there is a high linear correlation between the variables.When they are applied to the prediction of listed companies,the variables that can affect and predict financial distress are accurately filtered out.Among all the indicators,the indicators per share have the greatest influence while those of solvency have the weakest influence.Listed companies can assess the financial situation with the indicators screened by our algorithm and make an early warning of their possible financial distress in advance with higher precision.
基金Project supported by the National Natural Science Foundation of China(Grant No.11075184)the Knowledge Innovation Program of the Chinese Academy of Sciences(CAS)(Grant No.Y03RC21124)the CAS President’s International Fellowship Initiative Foundation(Grant No.2015VMA007)
文摘Laser-induced breakdown spectroscopy（LIBS） is a versatile tool for both qualitative and quantitative analysis.In this paper,LIBS combined with principal component analysis（PCA） and support vector machine（SVM） is applied to rock analysis.Fourteen emission lines including Fe,Mg,Ca,Al,Si,and Ti are selected as analysis lines.A good accuracy（91.38% for the real rock） is achieved by using SVM to analyze the spectroscopic peak area data which are processed by PCA.It can not only reduce the noise and dimensionality which contributes to improving the efficiency of the program,but also solve the problem of linear inseparability by combining PCA and SVM.By this method,the ability of LIBS to classify rock is validated.
文摘A method of applying support vector machine (SVM) in speech recognition was proposed, and a speech recognition system for mandarin digits was built up by SVMs. In the system, vectors were linearly extracted from speech feature sequence to make up time-aligned input patterns for SVM, and the decisions of several 2-class SVM classifiers were employed for constructing an N-class classifier. Four kinds of SVM kernel functions were compared in the experiments of speaker-independent speech recognition of mandarin digits. And the kernel of radial basis function has the highest accurate rate of 99.33%, which is better than that of the baseline system based on hidden Markov models (HMM) (97.08%). And the experiments also show that SVM can outperform HMM especially when the samples for learning were very limited.
基金Supported by the National Natural Science Foundation of China(60772056)the Postdoctoral Science Foundation of China(20070421094)
文摘An adaptive blind support vector machine equalizer（ABSVME） is presented in this paper.The method is based upon least square support vector machine（LSSVM）,and stems from signal feature reconstruction idea.By oversampling the output of a LSSVM equalizer and exploiting a reasonable decorrelation cost function design,the method achieves fine online channel tracing with Kumar express algorithm and static iterative learning algorithm incorporated.The method is verified through simulation and compared with other nonlinear equalizers.The results show that it provides excellent performance in nonlinear equalization and time-varying channel tracing.Although a constant module equalization algorithm requires that the signal has characteristic of constant module,this method has no such requirement.
文摘In order to solve the problems of small sample over-fitting and local minima when neural networks learn online, a novel method of predicting network bandwidth based on support vector machines(SVM) is proposed. The prediction and learning online will be completed by the proposed moving window learning algorithm(MWLA). The simulation research is done to validate the proposed method, which is compared with the method based on neural networks.
基金Sponsored by the National Natural Science Foundation of China(60472110)
文摘To improve the training speed of support vector machine （SVM）, a method called improved center distance ratio method （ICDRM） with determining thresholds automatically is presented here without reduce the identification rate. In this method border vectors are chosen from the given samples by comparing sample vectors with center distance ratio in advance. The number of training samples is reduced greatly and the training speed is improved. This method is used to the identification for license plate characters. Experimental resuhs show that the improved SVM method-ICDRM does well at identification rate and training speed.
