Equations(2)and(6)and the corresponding discussion in the paper[Chin.Phys.Lett.42,056301(2025)]have been corrected.These modiffcations do not affect the results derived in the paper.
To improve the classification performance of the kernel minimum squared error（ KMSE）, an enhanced KMSE algorithm（ EKMSE） is proposed. It redefines the regular objective function by introducing a novel class label ...To improve the classification performance of the kernel minimum squared error（ KMSE）, an enhanced KMSE algorithm（ EKMSE） is proposed. It redefines the regular objective function by introducing a novel class label definition, and the relative class label matrix can be adaptively adjusted to the kernel matrix.Compared with the common methods, the newobjective function can enlarge the distance between different classes, which therefore yields better recognition rates. In addition, an iteration parameter searching technique is adopted to improve the computational efficiency. The extensive experiments on FERET and GT face databases illustrate the feasibility and efficiency of the proposed EKMSE. It outperforms the original MSE, KMSE,some KMSE improvement methods, and even the sparse representation-based techniques in face recognition, such as collaborate representation classification（ CRC）.展开更多
The application of frequency distribution statistics to data provides objective means to assess the nature of the data distribution and viability of numerical models that are used to visualize and interpret data.Two c...The application of frequency distribution statistics to data provides objective means to assess the nature of the data distribution and viability of numerical models that are used to visualize and interpret data.Two commonly used tools are the kernel density estimation and reduced chi-squared statistic used in combination with a weighted mean.Due to the wide applicability of these tools,we present a Java-based computer application called KDX to facilitate the visualization of data and the utilization of these numerical tools.展开更多
As an important indicator parameter of fluid identification,fluid factor has always been a concern for scholars.However,when predicting Russell fluid factor or effective pore-fluid bulk modulus,it is necessary to intr...As an important indicator parameter of fluid identification,fluid factor has always been a concern for scholars.However,when predicting Russell fluid factor or effective pore-fluid bulk modulus,it is necessary to introduce a new rock skeleton parameter which is the dry-rock VP/VS ratio squared(DVRS).In the process of fluid factor calculation or inversion,the existing methods take this parameter as a static constant,which has been estimated in advance,and then apply it to the fluid factor calculation and inversion.The fluid identification analysis based on a portion of the Marmousi 2 model and numerical forward modeling test show that,taking the DVRS as a static constant will limit the identification ability of fluid factor and reduce the inversion accuracy.To solve the above problems,we proposed a new method to regard the DVRS as a dynamic variable varying with depth and lithology for the first time,then apply it to fluid factor calculation and inversion.Firstly,the exact Zoeppritz equations are rewritten into a new form containing the fluid factor and DVRS of upper and lower layers.Next,the new equations are applied to the four parameters simultaneous inversion based on the generalized nonlinear inversion(GNI)method.The testing results on a portion of the Marmousi 2 model and field data show that dynamic DVRS can significantly improve the fluid factor identification ability,effectively suppress illusion.Both synthetic and filed data tests also demonstrate that the GNI method based on Bayesian deterministic inversion(BDI)theory can successfully solve the above four parameter simultaneous inversion problem,and taking the dynamic DVRS as a target inversion parameter can effectively improve the inversion accuracy of fluid factor.All these results completely verified the feasibility and effectiveness of the proposed method.展开更多
In this article,we consider a discrete right-definite Sturm-Liouville problems with two squared eigenparameter-dependent boundary conditions.By constructing some new Lagrange-type identities and two fundamental functi...In this article,we consider a discrete right-definite Sturm-Liouville problems with two squared eigenparameter-dependent boundary conditions.By constructing some new Lagrange-type identities and two fundamental functions,we obtain not only the existence,the simplicity,and the interlacing properties of the real eigenvalues,but also the oscillation properties,orthogonality of the eigenfunctions,and the expansion theorem.Finally,we also give a computation scheme for computing eigenvalues and eigenfunctions of specific eigenvalue problems.展开更多
In this paper, the squared eigenfunction symmetries for the BTL and CTL hierarchies are expficitly constructed with the suitable modification of the ones for the TL hierarchy, by considering the BTL and GTL constraint...In this paper, the squared eigenfunction symmetries for the BTL and CTL hierarchies are expficitly constructed with the suitable modification of the ones for the TL hierarchy, by considering the BTL and GTL constraints. Also the connections with the corresponding additional symmetries are investigated： the squared eigenfunction symmetry generated by the wave function can be viewed as the generating function for the additional symmetries.展开更多
Analysis of stock recruitment (SR) data is most often done by fitting various SR relationship curves to the data. Fish population dynamics data often have stochastic variations and measurement errors, which usually re...Analysis of stock recruitment (SR) data is most often done by fitting various SR relationship curves to the data. Fish population dynamics data often have stochastic variations and measurement errors, which usually result in a biased regression analysis. This paper presents a robust regression method, least median of squared orthogonal distance (LMD), which is insensitive to abnormal values in the dependent and independent variables in a regression analysis. Outliers that have significantly different variance from the rest of the data can be identified in a residual analysis. Then, the least squares (LS) method is applied to the SR data with defined outliers being down weighted. The application of LMD and LMD based Reweighted Least Squares (RLS) method to simulated and real fisheries SR data is explored.展开更多
In this paper a square wavelet thresholding method is proposed and evaluated as compared to the other classical wavelet thresholding methods (like soft and hard). The main advantage of this work is to design and imple...In this paper a square wavelet thresholding method is proposed and evaluated as compared to the other classical wavelet thresholding methods (like soft and hard). The main advantage of this work is to design and implement a new wavelet thresholding method and evaluate it against other classical wavelet thresholding methods and hence search for the optimal wavelet mother function among the wide families with a suitable level of decomposition and followed by a novel thresholding method among the existing methods. This optimized method will be used to shrink the wavelet coefficients and yield an adequate compressed pressure signal prior to transmit it. While a comparison evaluation analysis is established, A new proposed procedure is used to compress a synthetic signal and obtain the optimal results through minimization the signal memory size and its transmission bandwidth. There are different performance indices to establish the comparison and evaluation process for signal compression;but the most well-known measuring scores are: NMSE, ESNR, and PDR. The obtained results showed the dominant of the square wavelet thresholding method against other methods using different measuring scores and hence the conclusion by the way for adopting this proposed novel wavelet thresholding method for 1D signal compression in future researches.展开更多
Three types of expression in the dark-soliton perturbation theory based on squared Jost solutions are invesgigaged in ghis paper. It is shown that there are three formally different results about the effects of pertur...Three types of expression in the dark-soliton perturbation theory based on squared Jost solutions are invesgigaged in ghis paper. It is shown that there are three formally different results about the effects of perturbagion on a dark soliton, and it is proved by means of a transformation between two integral variables that they are essentially equivalent.展开更多
With a special gauge transformation,the Lax pair of the derivative nonlinear Shcrdinger (DNLS) equation turns to depend on the squared parameter λ = k2instead of the usual spec-tral parameter k. By introducing a new ...With a special gauge transformation,the Lax pair of the derivative nonlinear Shcrdinger (DNLS) equation turns to depend on the squared parameter λ = k2instead of the usual spec-tral parameter k. By introducing a new direct product of Jost solu-tions,the complete Hamiltonian theory of the DNLS equation is constructed on the basis of the squared spectral parameter,which shows that the integrability completeness is still preserved. This result will be beneficial to the further study of the DNLS equation,such as the direct perturbation method.展开更多
The minimum squared Euclidean distance(MSED) of binary multi-h phase codes is presented. The signal segregation degree(SSD) has been put forward to determine MSED of multi-h phase codes. In order to maximize MSED, SSD...The minimum squared Euclidean distance(MSED) of binary multi-h phase codes is presented. The signal segregation degree(SSD) has been put forward to determine MSED of multi-h phase codes. In order to maximize MSED, SSD should be as large as possible. The necessary and sufficient conditions of maximizing SSD are derived. Finally, SSD and the exact formulae for MSED of binary 2-h phase codes are also presented.展开更多
We investigate the linear cosmological perturbations in the context of the so-called energy-momentum squared gravity(EMSG)theory.Recent research shows that the EMSG theory can reproduce a viable background cosmologica...We investigate the linear cosmological perturbations in the context of the so-called energy-momentum squared gravity(EMSG)theory.Recent research shows that the EMSG theory can reproduce a viable background cosmological evolution comparable toΛCDM,whereas the matter-dominated era exhibits slight distinctions.In this paper,we focus on power-law EMSG models and derive the equations for the linear cosmological perturbations.We explore the propagation of the gravitational wave(GW)and the growth of matter density perturbation at the first order,and we estimate the model parameters from the simulated GW and observed redshift space distortion data.Our analysis reveals that the model parameters should be small and positive in the confidence interval,which indicates that the theory agrees closely with the observational data and can be considered an alternative to the standard cosmological model.展开更多
This paper deals with the numerical solution of the large-scale Stein and discrete-time Lyapunov matrix equations.Based on the global Arnoldi process and the squared Smith iteration,we propose a low-rank global Krylov...This paper deals with the numerical solution of the large-scale Stein and discrete-time Lyapunov matrix equations.Based on the global Arnoldi process and the squared Smith iteration,we propose a low-rank global Krylov squared Smith method for solving large-scale Stein and discrete-time Lyapunov matrix equations,and estimate the upper bound of the error and the residual of the approximate solutions for the matrix equations.Moreover,we discuss the restarting of the low-rank global Krylov squared Smith method and provide some numerical experiments to show the efficiency of the proposed method.展开更多
In the paper, a novel practical approach to construct a composite indicator （CI） is pro- posed. The key idea is to decide the weights of sub-indicators in constructing a composite indicator by maximizing the sum of ...In the paper, a novel practical approach to construct a composite indicator （CI） is pro- posed. The key idea is to decide the weights of sub-indicators in constructing a composite indicator by maximizing the sum of squared correlations between the CI and sub-indicators. The CI obtained in this fashion has the maximum sum of squared correlations among all linear estimators. In addition, the simple, exact and explicit solutions of weights are proposed under the condition of non-negative irreducible matrix. Moreover, under this particular condition, the proposed method will become the principal component analysis. For illustration purpose, the proposed novel approach is utilized to cal- culate Sustainable Energy Index and Human Development Index which are two often-used cases to compare models in the literatures. The results the methods of Hatefi, et al. （2010） and Zhou all sub-indicators＇ correlations. show that the power of the proposed method outweigh et al. （2007） in terms of the sum of absolute values of展开更多
Minimum squared error(MSE)algorithm is one of the classical pattern recognition and regression analysis methods,whose objective is to minimize the squared error summation between the output of linear function and the ...Minimum squared error(MSE)algorithm is one of the classical pattern recognition and regression analysis methods,whose objective is to minimize the squared error summation between the output of linear function and the desired output.In this paper,the MSE algorithm is modified by using kernel functions satisfying the Mercer condition and regularization technique;and the nonlinear MSE algorithms based on kernels and regularization term,that is,the regularized kernel forms of MSE algorithm,are proposed.Their objective functions include the squared error summation between the output of nonlinear function based on kernels and the desired output and a proper regularization term.The regularization technique can handle ill-posed problems,reduce the solution space,and control the generalization.Three squared regularization terms are utilized in this paper.In accordance with the probabilistic interpretation of regularization terms,the difference among three regularization terms is given in detail.The synthetic and real data are used to analyze the algorithm performance.展开更多
In the realm of survey data analysis,encountering substantial variance relative to bias is a common occurrence.In this study,we present an innovative strategy to tackle this issue by introducing slightly biased varian...In the realm of survey data analysis,encountering substantial variance relative to bias is a common occurrence.In this study,we present an innovative strategy to tackle this issue by introducing slightly biased variance estimators.These estimators incorporate a constant c within the range of 0 to 1,which is determined through the minimization of Mean Squared Error(MSE)for c×(variance estimator).This research builds upon the foundation laid by Kourouklis(2012,A new estimator of the variance based on minimizing mean squared error.The American Statistician,66(4),234–236)and extends their work into the domain of survey sampling.Extensive simulation studies are conducted to illustrate the superior performance of the adjusted variance estimators when compared to standard variance estimators,particularly in terms of MSE.These findings underscore the efficacy of our proposed approach in enhancing the precision of variance estimation within the context of survey data analysis.展开更多
For classification problems,the traditional least squares twin support vector machine(LSTSVM)generates two nonparallel hyperplanes directly by solving two systems of linear equations instead of a pair of quadratic pro...For classification problems,the traditional least squares twin support vector machine(LSTSVM)generates two nonparallel hyperplanes directly by solving two systems of linear equations instead of a pair of quadratic programming problems(QPPs),which makes LSTSVM much faster than the original TSVM.But the standard LSTSVM adopting quadratic loss measured by the minimal distance is sensitive to noise and unstable to re-sampling.To overcome this problem,the expectile distance is taken into consideration to measure the margin between classes and LSTSVM with asymmetric squared loss(aLSTSVM)is proposed.Compared to the original LSTSVM with the quadratic loss,the proposed aLSTSVM not only has comparable computational accuracy,but also performs good properties such as noise insensitivity,scatter minimization and re-sampling stability.Numerical experiments on synthetic datasets,normally distributed clustered(NDC)datasets and University of California,Irvine(UCI)datasets with different noises confirm the great performance and validity of our proposed algorithm.展开更多
文摘Equations(2)and(6)and the corresponding discussion in the paper[Chin.Phys.Lett.42,056301(2025)]have been corrected.These modiffcations do not affect the results derived in the paper.
基金The Priority Academic Program Development of Jiangsu Higher Education Institutions(PAPD)the National Natural Science Foundation of China(No.61572258,61103141,51405241)+1 种基金the Natural Science Foundation of Jiangsu Province(No.BK20151530)Overseas Training Programs for Outstanding Young Scholars of Universities in Jiangsu Province
文摘To improve the classification performance of the kernel minimum squared error（ KMSE）, an enhanced KMSE algorithm（ EKMSE） is proposed. It redefines the regular objective function by introducing a novel class label definition, and the relative class label matrix can be adaptively adjusted to the kernel matrix.Compared with the common methods, the newobjective function can enlarge the distance between different classes, which therefore yields better recognition rates. In addition, an iteration parameter searching technique is adopted to improve the computational efficiency. The extensive experiments on FERET and GT face databases illustrate the feasibility and efficiency of the proposed EKMSE. It outperforms the original MSE, KMSE,some KMSE improvement methods, and even the sparse representation-based techniques in face recognition, such as collaborate representation classification（ CRC）.
文摘The application of frequency distribution statistics to data provides objective means to assess the nature of the data distribution and viability of numerical models that are used to visualize and interpret data.Two commonly used tools are the kernel density estimation and reduced chi-squared statistic used in combination with a weighted mean.Due to the wide applicability of these tools,we present a Java-based computer application called KDX to facilitate the visualization of data and the utilization of these numerical tools.
基金the National Natural Science Foundation of China(41904116,41874156,42074167 and 42204135)the Natural Science Foundation of Hunan Province(2020JJ5168)the China Postdoctoral Science Foundation(2021M703629)for their funding of this research.
文摘As an important indicator parameter of fluid identification,fluid factor has always been a concern for scholars.However,when predicting Russell fluid factor or effective pore-fluid bulk modulus,it is necessary to introduce a new rock skeleton parameter which is the dry-rock VP/VS ratio squared(DVRS).In the process of fluid factor calculation or inversion,the existing methods take this parameter as a static constant,which has been estimated in advance,and then apply it to the fluid factor calculation and inversion.The fluid identification analysis based on a portion of the Marmousi 2 model and numerical forward modeling test show that,taking the DVRS as a static constant will limit the identification ability of fluid factor and reduce the inversion accuracy.To solve the above problems,we proposed a new method to regard the DVRS as a dynamic variable varying with depth and lithology for the first time,then apply it to fluid factor calculation and inversion.Firstly,the exact Zoeppritz equations are rewritten into a new form containing the fluid factor and DVRS of upper and lower layers.Next,the new equations are applied to the four parameters simultaneous inversion based on the generalized nonlinear inversion(GNI)method.The testing results on a portion of the Marmousi 2 model and field data show that dynamic DVRS can significantly improve the fluid factor identification ability,effectively suppress illusion.Both synthetic and filed data tests also demonstrate that the GNI method based on Bayesian deterministic inversion(BDI)theory can successfully solve the above four parameter simultaneous inversion problem,and taking the dynamic DVRS as a target inversion parameter can effectively improve the inversion accuracy of fluid factor.All these results completely verified the feasibility and effectiveness of the proposed method.
基金The authors are supported by National Natural Sciences Foundation of China(11961060,11671322)the Key Project of Natural Sciences Foundation of Gansu Province(18JR3RA084).
文摘In this article,we consider a discrete right-definite Sturm-Liouville problems with two squared eigenparameter-dependent boundary conditions.By constructing some new Lagrange-type identities and two fundamental functions,we obtain not only the existence,the simplicity,and the interlacing properties of the real eigenvalues,but also the oscillation properties,orthogonality of the eigenfunctions,and the expansion theorem.Finally,we also give a computation scheme for computing eigenvalues and eigenfunctions of specific eigenvalue problems.
基金Supported by National Natural Science Foundation of China under Grant No. 11226196the Fundamental Research Funds for the Central Universities under Grant No. 2012QNA45
文摘In this paper, the squared eigenfunction symmetries for the BTL and CTL hierarchies are expficitly constructed with the suitable modification of the ones for the TL hierarchy, by considering the BTL and GTL constraints. Also the connections with the corresponding additional symmetries are investigated： the squared eigenfunction symmetry generated by the wave function can be viewed as the generating function for the additional symmetries.
文摘Analysis of stock recruitment (SR) data is most often done by fitting various SR relationship curves to the data. Fish population dynamics data often have stochastic variations and measurement errors, which usually result in a biased regression analysis. This paper presents a robust regression method, least median of squared orthogonal distance (LMD), which is insensitive to abnormal values in the dependent and independent variables in a regression analysis. Outliers that have significantly different variance from the rest of the data can be identified in a residual analysis. Then, the least squares (LS) method is applied to the SR data with defined outliers being down weighted. The application of LMD and LMD based Reweighted Least Squares (RLS) method to simulated and real fisheries SR data is explored.
文摘In this paper a square wavelet thresholding method is proposed and evaluated as compared to the other classical wavelet thresholding methods (like soft and hard). The main advantage of this work is to design and implement a new wavelet thresholding method and evaluate it against other classical wavelet thresholding methods and hence search for the optimal wavelet mother function among the wide families with a suitable level of decomposition and followed by a novel thresholding method among the existing methods. This optimized method will be used to shrink the wavelet coefficients and yield an adequate compressed pressure signal prior to transmit it. While a comparison evaluation analysis is established, A new proposed procedure is used to compress a synthetic signal and obtain the optimal results through minimization the signal memory size and its transmission bandwidth. There are different performance indices to establish the comparison and evaluation process for signal compression;but the most well-known measuring scores are: NMSE, ESNR, and PDR. The obtained results showed the dominant of the square wavelet thresholding method against other methods using different measuring scores and hence the conclusion by the way for adopting this proposed novel wavelet thresholding method for 1D signal compression in future researches.
基金The project supported by National Natural Science Foundation of China under Grant No. 10375022 and the Scientific Research Fund of the Education Department of Hunan Province of China under Grant No. 05C414
文摘Three types of expression in the dark-soliton perturbation theory based on squared Jost solutions are invesgigaged in ghis paper. It is shown that there are three formally different results about the effects of perturbagion on a dark soliton, and it is proved by means of a transformation between two integral variables that they are essentially equivalent.
基金Supported by the National Natural Science Foundation of China (10705022)
文摘With a special gauge transformation,the Lax pair of the derivative nonlinear Shcrdinger (DNLS) equation turns to depend on the squared parameter λ = k2instead of the usual spec-tral parameter k. By introducing a new direct product of Jost solu-tions,the complete Hamiltonian theory of the DNLS equation is constructed on the basis of the squared spectral parameter,which shows that the integrability completeness is still preserved. This result will be beneficial to the further study of the DNLS equation,such as the direct perturbation method.
文摘The minimum squared Euclidean distance(MSED) of binary multi-h phase codes is presented. The signal segregation degree(SSD) has been put forward to determine MSED of multi-h phase codes. In order to maximize MSED, SSD should be as large as possible. The necessary and sufficient conditions of maximizing SSD are derived. Finally, SSD and the exact formulae for MSED of binary 2-h phase codes are also presented.
基金Supported by the National SKA Program of China(2022SKA0110200,2022SKA0110203)the National Natural Science Foundation of China(12473001,11975072,12247101)the National 111 Project(B16009)。
文摘We investigate the linear cosmological perturbations in the context of the so-called energy-momentum squared gravity(EMSG)theory.Recent research shows that the EMSG theory can reproduce a viable background cosmological evolution comparable toΛCDM,whereas the matter-dominated era exhibits slight distinctions.In this paper,we focus on power-law EMSG models and derive the equations for the linear cosmological perturbations.We explore the propagation of the gravitational wave(GW)and the growth of matter density perturbation at the first order,and we estimate the model parameters from the simulated GW and observed redshift space distortion data.Our analysis reveals that the model parameters should be small and positive in the confidence interval,which indicates that the theory agrees closely with the observational data and can be considered an alternative to the standard cosmological model.
基金supported by the National Natural Science Foundation of China under grant No.11571171.
文摘This paper deals with the numerical solution of the large-scale Stein and discrete-time Lyapunov matrix equations.Based on the global Arnoldi process and the squared Smith iteration,we propose a low-rank global Krylov squared Smith method for solving large-scale Stein and discrete-time Lyapunov matrix equations,and estimate the upper bound of the error and the residual of the approximate solutions for the matrix equations.Moreover,we discuss the restarting of the low-rank global Krylov squared Smith method and provide some numerical experiments to show the efficiency of the proposed method.
基金supported by National Natural Science Foundation of China under Grant No.71003100the Fundamental Research Funds for the Central Universitiesthe Research Funds of Renmin University of China under Grant No.11XNK027
文摘In the paper, a novel practical approach to construct a composite indicator （CI） is pro- posed. The key idea is to decide the weights of sub-indicators in constructing a composite indicator by maximizing the sum of squared correlations between the CI and sub-indicators. The CI obtained in this fashion has the maximum sum of squared correlations among all linear estimators. In addition, the simple, exact and explicit solutions of weights are proposed under the condition of non-negative irreducible matrix. Moreover, under this particular condition, the proposed method will become the principal component analysis. For illustration purpose, the proposed novel approach is utilized to cal- culate Sustainable Energy Index and Human Development Index which are two often-used cases to compare models in the literatures. The results the methods of Hatefi, et al. （2010） and Zhou all sub-indicators＇ correlations. show that the power of the proposed method outweigh et al. （2007） in terms of the sum of absolute values of
基金supported by the National Natural Science Foundation of China (No.60275007 and No.698885004).
文摘Minimum squared error(MSE)algorithm is one of the classical pattern recognition and regression analysis methods,whose objective is to minimize the squared error summation between the output of linear function and the desired output.In this paper,the MSE algorithm is modified by using kernel functions satisfying the Mercer condition and regularization technique;and the nonlinear MSE algorithms based on kernels and regularization term,that is,the regularized kernel forms of MSE algorithm,are proposed.Their objective functions include the squared error summation between the output of nonlinear function based on kernels and the desired output and a proper regularization term.The regularization technique can handle ill-posed problems,reduce the solution space,and control the generalization.Three squared regularization terms are utilized in this paper.In accordance with the probabilistic interpretation of regularization terms,the difference among three regularization terms is given in detail.The synthetic and real data are used to analyze the algorithm performance.
文摘In the realm of survey data analysis,encountering substantial variance relative to bias is a common occurrence.In this study,we present an innovative strategy to tackle this issue by introducing slightly biased variance estimators.These estimators incorporate a constant c within the range of 0 to 1,which is determined through the minimization of Mean Squared Error(MSE)for c×(variance estimator).This research builds upon the foundation laid by Kourouklis(2012,A new estimator of the variance based on minimizing mean squared error.The American Statistician,66(4),234–236)and extends their work into the domain of survey sampling.Extensive simulation studies are conducted to illustrate the superior performance of the adjusted variance estimators when compared to standard variance estimators,particularly in terms of MSE.These findings underscore the efficacy of our proposed approach in enhancing the precision of variance estimation within the context of survey data analysis.
基金supported in part by the National Natural Science Foundation of China(51875457)Natural Science Foundation of Shaanxi Province of China(2021JQ-701)+1 种基金the Key Research Project of Shaanxi Province(2022GY-050,2022GY-028)Xi’an Science and Technology Plan Project(2020KJRC0109)。
文摘For classification problems,the traditional least squares twin support vector machine(LSTSVM)generates two nonparallel hyperplanes directly by solving two systems of linear equations instead of a pair of quadratic programming problems(QPPs),which makes LSTSVM much faster than the original TSVM.But the standard LSTSVM adopting quadratic loss measured by the minimal distance is sensitive to noise and unstable to re-sampling.To overcome this problem,the expectile distance is taken into consideration to measure the margin between classes and LSTSVM with asymmetric squared loss(aLSTSVM)is proposed.Compared to the original LSTSVM with the quadratic loss,the proposed aLSTSVM not only has comparable computational accuracy,but also performs good properties such as noise insensitivity,scatter minimization and re-sampling stability.Numerical experiments on synthetic datasets,normally distributed clustered(NDC)datasets and University of California,Irvine(UCI)datasets with different noises confirm the great performance and validity of our proposed algorithm.
