We study the quasi-random choice method (QRCM) for the Liouville equation of ge- ometrical optics with discontinuous locM wave speed. This equation arises in the phase space computation of high frequency waves throu...
详细信息
We study the quasi-random choice method (QRCM) for the Liouville equation of ge- ometrical optics with discontinuous locM wave speed. This equation arises in the phase space computation of high frequency waves through interfaces, where waves undergo partial transmissions and reflections. The numerical challenges include interface, contact discon- tinuities, and measure-valued solutions. The so-called QRCM is a random choice method based on quasi-random sampling (a deterministic alternative to random sampling). The method not only is viscosity-free but also provides faster convergence rate. Therefore, it is appealing for the prob!em under study which is indeed a Hamiltonian flow. Our analy- sis and computational results show that the QRCM 1) is almost first-order accurate even with the aforementioned discontinuities; 2) gives sharp resolutions for all discontinuities encountered in the problem; and 3) for measure-valued solutions, does not need the level set decomposition for finite difference/volume methods with numerical viscosities.
The PADS (Process Algebra for Demand and Supply) framework is an approach to model resource demand and supply for the formal analysis of hierarchical scheduling. Inspired by the demand relation in PADS, we propose a w...
详细信息
This paper studied subspace properties of the Celis–Dennis–Tapia(CDT)subproblem that arises in some trust-region algorithms for equality constrained opti*** analysis is an extension of that presented by Wang and Yu...
详细信息
This paper studied subspace properties of the Celis–Dennis–Tapia(CDT)subproblem that arises in some trust-region algorithms for equality constrained opti*** analysis is an extension of that presented by Wang and Yuan(***.104:241–269,2006)for the standard trust-region *** suitable conditions,it is shown that the trial step obtained from the CDT subproblem is in the subspace spanned by all the gradient vectors of the objective function and of the constraints computed until the current *** on this observation,a subspace version of the Powell–Yuan trust-region algorithm is proposed for equality constrained optimization problems where the number of constraints is much lower than the number of variables. The convergence analysis is given and numerical results arealso reported.
作者:
Zhenli Xu[]Department of Mathematics
Institute of Natural Sciences and Ministry of Education Key Laboratory in Scientific and Engineering Computing Shanghai Jiao Tong University Shanghai 200240 China
Electrostatic polarization is important in many nano- and micro-scale physical systems such as colloidal suspensions, biopolymers, and nanomaterials assembly. The calculation of polarization potential requires an effi...
详细信息
Electrostatic polarization is important in many nano- and micro-scale physical systems such as colloidal suspensions, biopolymers, and nanomaterials assembly. The calculation of polarization potential requires an efficient algorithm for solving 3D Poisson's equation. We have developed a useful image charge method to rapid evaluation of the Green's function of the Poisson's equation in the presence of spherical dielectric discontinuities. This paper presents an extensive study of this method by giving a convergence analysis and developing a coarse-graining algorithm. The use of the coarse graining could reduce the number of image charges to around a dozen, by 1–2 orders of magnitude. We use the algorithm to investigate the interaction force between likely charged spheres in different dielectric environments. We find the size and charge asymmetry leads to an attraction between like charges, in agreement with existing results. Furthermore, we study three-body interactions and find, in the presence of an external interface, that the interaction force depends on the curvature of the interface and performs a nonmonotonic electrostatic force.
In statistical word alignment for machine translation, function words usually cause poor aligning performance because they do not have clear correspondence between different languages. This paper proposes a novel appr...
详细信息
ISBN:
(纸本)9781577356332
In statistical word alignment for machine translation, function words usually cause poor aligning performance because they do not have clear correspondence between different languages. This paper proposes a novel approach to improve word alignment by pruning alignments of function words from an existing alignment model with high precision and recall. Based on monolingual and bilingual frequency characteristics, a language-independent function word recognition algorithm is first proposed. Then a group of carefully defined syntactic structures combined with content word alignments are used for further function word alignment pruning. The experimental results show that the proposed approach improves both the quality of word alignment and the performance of statistical machine translation on Chinese-to-English, Germanto- English and French-to-English language pairs.
Difference equations or discrete systems are mathematical models of various fields such as physics, chemistry, biology, and economics and have been subjects of extensive study of both pure mathematicians and applied m...
详细信息
Difference equations or discrete systems are mathematical models of various fields such as physics, chemistry, biology, and economics and have been subjects of extensive study of both pure mathematicians and applied mathematicians. Through its interaction with modern integrable systems, the theory of difference equations is enriched greatly and has been undergoing a rapid development. SIDE-10, the tenth of a series of biennial conferences devoted to Symmetries and Integrability of Difference Equations and related topics, was held during 10-16 June, 2012 at Ningbo, China. It was sponsored and supported by the National Natural Science Foundation of China, Ningbo Association of Science and Technology, Ningbo University, Academy of Mathematics and Systems Science of Chinese Academy of Sciences, China University of Mining and Technology (Beijing), Tsinghua University, and Shanghai University. The conference attracted over 100 participants from more than a dozen of countries. During the conference, 44 contributed talks were arranged and the topics covered by the meeting include
The global boundness and existence are presented for the kind of the Rosseland equation with a general growth condition. A linearized map in a closed convex set is defined. The image set is precompact, and thus a fixe...
详细信息
The global boundness and existence are presented for the kind of the Rosseland equation with a general growth condition. A linearized map in a closed convex set is defined. The image set is precompact, and thus a fixed point exists. A multi-scale expansion method is used to obtain the homogenized equation. This equation satisfies a similar growth condition.
Neural network language models, or continuous-space language models (CSLMs), have been shown to improve the performance of statistical machine translation (SMT) when they are used for reranking n-best translations. Ho...
详细信息
作者:
Mingfen LiJie JiaYe LiuDepartment of Rehabilitation
Huashan Hospital Fudan University MOE-Microsoft Key Laboratory for Intelligent Computing and Intelligent Systems Department of Computer Science and Engineering Shanghai Jiao Tong University China
Word segmentation has been shown helpful for Chinese-to-English machine translation (MT), yet the way different segmentation strategies affect MT is poorly understood. In this paper, we focus on comparing different se...
详细信息
暂无评论