考虑将特征选择集成到支持向量机分类器中,提出集成特征选择的最优化支持向量机分类器——FS-SDPSVM(Feature Selection in Semi-definite Program for Support Vector Machine)。该模型将每个特征分别在核空间中做特征映射,然后通过参...考虑将特征选择集成到支持向量机分类器中,提出集成特征选择的最优化支持向量机分类器——FS-SDPSVM(Feature Selection in Semi-definite Program for Support Vector Machine)。该模型将每个特征分别在核空间中做特征映射,然后通过参数组合构成新的核矩阵,将特征选择过程与机器分类过程统一在一个优化目标下,同时达到特征选择与分类最优。在特征筛选方面,根据模型参数提出用于特征筛选的特征支持度和特征贡献度,通过控制二者的上下限可以在最优分类和最少特征之间灵活取舍。实证中分别将最优分类(FS-SDP-SVM1)和最少特征(FS-SDPSVM2)两类集成化特征选择算法与Relief-F、SFS、SBS算法在UCI机器学习数据和人造数据中进行对比实验。结果表明,提出的FS-SDP-SVM算法在保持较好泛化能力的基础上,在多数实验数据集中实现了最大分类准确率或最少特征数量;在人工数据中,该方法可以准确地选出真正的特征,去除噪声特征。展开更多
We study the energy level statistics of the SO(5) limit of super-symmetry U(6/4) in odd-A nucleus using the interacting boson-fermion model. The nearest neighbor spacing distribution (NSD) and the spectral rigidity (...We study the energy level statistics of the SO(5) limit of super-symmetry U(6/4) in odd-A nucleus using the interacting boson-fermion model. The nearest neighbor spacing distribution (NSD) and the spectral rigidity (△3)are investigated, and the factors that affect the properties of level statistics are also discussed. The results show that the boson number N is a dominant factor. If N is small, both the interaction strengths of subgroups SOB(5) and SOBF(5)and the spin play important roles in the energy level statistics, however, along with the increase of N, the statistics distribution would tend to be in Poisson form.展开更多
Prequantale quotient is defined in terms of prequantic nucleus, some equivalent propositions of which is obtained. Also, prequantale quotient is proved to coincide with the quotient object in PQuant up to isomorphism....Prequantale quotient is defined in terms of prequantic nucleus, some equivalent propositions of which is obtained. Also, prequantale quotient is proved to coincide with the quotient object in PQuant up to isomorphism. At last, the concrete structure of the largest spatial frame quotient in TPQuant is given.展开更多
This paper is a short revisit to Kuo-Brown effective interaction derived from the Hamada-Johnston nucleon-nucleon potential, done by Gerry Brown and Tom Kuo. This effective interaction, derived in year 1966, is the fi...This paper is a short revisit to Kuo-Brown effective interaction derived from the Hamada-Johnston nucleon-nucleon potential, done by Gerry Brown and Tom Kuo. This effective interaction, derived in year 1966, is the first attempt to describe nuclear structure properties from the free nucleon-nucleon potential. Nowadays much progress has been achieved for the effective interactions in shell model. We would compare the effective interactions obtained in the 1966 paper with up-to-date shell-model interactions in sd-shell and pf-shell model space. Recent knowledge of effective interactions on nuclear structure, can also be traced in the KuoBrown effective interaction, i.e., the universal roles of central and tensor forces, which reminds us that such discovery should be noticed much earlier.展开更多
The authors define the equi-nuclearity of uniform Roe algebras of a family of metric spaces. For a discrete metric space X with bounded geometry which is covered by a family of subspaces {Xi}i=1^∞, if {C^＊（Xi）}i=1...The authors define the equi-nuclearity of uniform Roe algebras of a family of metric spaces. For a discrete metric space X with bounded geometry which is covered by a family of subspaces {Xi}i=1^∞, if {C^＊（Xi）}i=1^∞ are equi-nuclear and under some proper gluing conditions, it is proved that C＊（X） is nuclear. Furthermore, it is claimed that in general, the coarse Roe algebra C^＊ （X） is not nuclear.展开更多
Semi-supervised learning is an emerging computational paradigm for machine learning,that aims to make better use of large amounts of inexpensive unlabeled data to improve the learning performance.While various methods...Semi-supervised learning is an emerging computational paradigm for machine learning,that aims to make better use of large amounts of inexpensive unlabeled data to improve the learning performance.While various methods have been proposed based on different intuitions,the crucial issue of generalization performance is still poorly understood.In this paper,we investigate the convergence property of the Laplacian regularized least squares regression,a semi-supervised learning algorithm based on manifold regularization.Moreover,the improvement of error bounds in terms of the number of labeled and unlabeled data is presented for the first time as far as we know.The convergence rate depends on the approximation property and the capacity of the reproducing kernel Hilbert space measured by covering numbers.Some new techniques are exploited for the analysis since an extra regularizer is introduced.展开更多
CSES(China Seismo-Electromagnetic Satellite) is a mission developed by CNSA(Chinese National Space Administration) and ASI(Italian Space Agency), to investigate the near-Earth electromagnetic, plasma and particle envi...CSES(China Seismo-Electromagnetic Satellite) is a mission developed by CNSA(Chinese National Space Administration) and ASI(Italian Space Agency), to investigate the near-Earth electromagnetic, plasma and particle environment, for studying the seismo-associated disturbances in the ionosphere-magnetosphere transition zone. The anthropogenic and electromagnetic noise,as well as the natural non-seismic electromagnetic emissions is mainly due to tropospheric activity. In particular, the mission aims to confirming the existence of possible temporal correlations between the occurrence of earthquakes for medium and strong magnitude and the observation in space of electromagnetic perturbations, plasma variations and precipitation of bursts with highenergy charged particles from the inner Van Allen belt. In this framework, the high energy particle detector(HEPD) of the CSES mission has been developed by the Italian LIMADOU Collaboration. HEPD is an advanced detector based on a tower of scintillators and a silicon tracker that provides good energy and angular resolution and a wide angular acceptance, for electrons of 3–100 Me V, protons of 30–200 Me V and light nuclei up to the oxygen. CSES satellite has been launched on February 2^(nd), 2018 from the Jiuquan Satellite Launch Center(China).展开更多
文摘考虑将特征选择集成到支持向量机分类器中,提出集成特征选择的最优化支持向量机分类器——FS-SDPSVM(Feature Selection in Semi-definite Program for Support Vector Machine)。该模型将每个特征分别在核空间中做特征映射,然后通过参数组合构成新的核矩阵,将特征选择过程与机器分类过程统一在一个优化目标下,同时达到特征选择与分类最优。在特征筛选方面,根据模型参数提出用于特征筛选的特征支持度和特征贡献度,通过控制二者的上下限可以在最优分类和最少特征之间灵活取舍。实证中分别将最优分类(FS-SDP-SVM1)和最少特征(FS-SDPSVM2)两类集成化特征选择算法与Relief-F、SFS、SBS算法在UCI机器学习数据和人造数据中进行对比实验。结果表明,提出的FS-SDP-SVM算法在保持较好泛化能力的基础上,在多数实验数据集中实现了最大分类准确率或最少特征数量;在人工数据中,该方法可以准确地选出真正的特征,去除噪声特征。
文摘We study the energy level statistics of the SO(5) limit of super-symmetry U(6/4) in odd-A nucleus using the interacting boson-fermion model. The nearest neighbor spacing distribution (NSD) and the spectral rigidity (△3)are investigated, and the factors that affect the properties of level statistics are also discussed. The results show that the boson number N is a dominant factor. If N is small, both the interaction strengths of subgroups SOB(5) and SOBF(5)and the spin play important roles in the energy level statistics, however, along with the increase of N, the statistics distribution would tend to be in Poisson form.
基金Supported by the National Natural Science Foundation of China(10471083)
文摘Prequantale quotient is defined in terms of prequantic nucleus, some equivalent propositions of which is obtained. Also, prequantale quotient is proved to coincide with the quotient object in PQuant up to isomorphism. At last, the concrete structure of the largest spatial frame quotient in TPQuant is given.
基金supported by the National Natural Science Foundation of China(Grant Nos.11275067 and 11447109)the support from the Helmholtz Association(HGF)through the Nuclear Astrophysics Virtual Institute(VH-VI-417)
文摘This paper is a short revisit to Kuo-Brown effective interaction derived from the Hamada-Johnston nucleon-nucleon potential, done by Gerry Brown and Tom Kuo. This effective interaction, derived in year 1966, is the first attempt to describe nuclear structure properties from the free nucleon-nucleon potential. Nowadays much progress has been achieved for the effective interactions in shell model. We would compare the effective interactions obtained in the 1966 paper with up-to-date shell-model interactions in sd-shell and pf-shell model space. Recent knowledge of effective interactions on nuclear structure, can also be traced in the KuoBrown effective interaction, i.e., the universal roles of central and tensor forces, which reminds us that such discovery should be noticed much earlier.
基金supported by the National Natural Science Foundation of China(Nos.10731020,10971023)the Shu Guang Project of Shanghai Municipal Education Commission and Shanghai Education DepartmentFoundation(No.07SG38)the Foundation of the Ministry of Education of China
文摘The authors define the equi-nuclearity of uniform Roe algebras of a family of metric spaces. For a discrete metric space X with bounded geometry which is covered by a family of subspaces {Xi}i=1^∞, if {C^＊（Xi）}i=1^∞ are equi-nuclear and under some proper gluing conditions, it is proved that C＊（X） is nuclear. Furthermore, it is claimed that in general, the coarse Roe algebra C^＊ （X） is not nuclear.
基金supported by National Natural Science Foundation of China (Grant Nos.11171014 and 11101024)National Basic Research Program of China (973 Project) (Grant No. 2010CB731900)
文摘Semi-supervised learning is an emerging computational paradigm for machine learning,that aims to make better use of large amounts of inexpensive unlabeled data to improve the learning performance.While various methods have been proposed based on different intuitions,the crucial issue of generalization performance is still poorly understood.In this paper,we investigate the convergence property of the Laplacian regularized least squares regression,a semi-supervised learning algorithm based on manifold regularization.Moreover,the improvement of error bounds in terms of the number of labeled and unlabeled data is presented for the first time as far as we know.The convergence rate depends on the approximation property and the capacity of the reproducing kernel Hilbert space measured by covering numbers.Some new techniques are exploited for the analysis since an extra regularizer is introduced.
基金supported by the Italian Space Agency in the framework of the“Accordo Attuativo n.2016-16-H0 Progetto Limadou Fase E/Scienza”(CUP F12F1600011005)
文摘CSES(China Seismo-Electromagnetic Satellite) is a mission developed by CNSA(Chinese National Space Administration) and ASI(Italian Space Agency), to investigate the near-Earth electromagnetic, plasma and particle environment, for studying the seismo-associated disturbances in the ionosphere-magnetosphere transition zone. The anthropogenic and electromagnetic noise,as well as the natural non-seismic electromagnetic emissions is mainly due to tropospheric activity. In particular, the mission aims to confirming the existence of possible temporal correlations between the occurrence of earthquakes for medium and strong magnitude and the observation in space of electromagnetic perturbations, plasma variations and precipitation of bursts with highenergy charged particles from the inner Van Allen belt. In this framework, the high energy particle detector(HEPD) of the CSES mission has been developed by the Italian LIMADOU Collaboration. HEPD is an advanced detector based on a tower of scintillators and a silicon tracker that provides good energy and angular resolution and a wide angular acceptance, for electrons of 3–100 Me V, protons of 30–200 Me V and light nuclei up to the oxygen. CSES satellite has been launched on February 2^(nd), 2018 from the Jiuquan Satellite Launch Center(China).
