检索结果-南通市图书馆

Derivation of Expanded Isospectral-Nonisospectral Integrable Hierarchies via the Column-vector Loop Algebra

在线全文

学校读者我要写书评

暂无评论

Acta Mathematicae Applicatae Sinica 2024年第3期40卷 778-800页

作者： hai-feng WANG Yu-feng zhang School of Science Jimei UniversityXiamen 361021China School of Mathematics China University of Mining and TechnologyXuzhou 221116China

A scheme for generating nonisospectral integrable hierarchies is *** on the method,we deduce a nonisospectral hierarchy of soliton equations by considering a linear spectral *** follows that the corresponding expanded isospectral and nonisospectral integrable hierarchies are deduced based on a 6 dimensional complex linear space ■.By reducing these integrable hierarchies,we obtain the expanded isospectral and nonisospectral derivative nonlinear Schr?dinger *** using the trace identity,the biHamiltonian structure of these two hierarchies are also ***,some symmetries and conserved quantities of the resulting hierarchy are discussed.

关键词： expanded isospectral-nonisospectral integrable hierarchies column-vector loop algebra bi-Hamiltonian structure symmetry

What and where: A context-based recommendation system for object insertion

在线全文

学校读者我要写书评

暂无评论

Computational Visual Media 2020年第1期6卷 79-93页

作者： song-hai zhang Zheng-Ping Zhou Bin Liu Xi Dong Peter Hall Tsinghua University Beijing 100084China Beijing National Research Center for Information Science and Technology(BNRist) Tsinghua UniversityBeijing 100084China Department of Computer Science Media Technology Research Center University of BathBath BA27AYUK

We propose a novel problem revolving around two tasks:(i)given a scene,recommend objects to insert,and(ii)given an object category,retrieve suitable background scenes.A bounding box for the inserted object is predicted in both tasks,which helps downstream applications such as semiautomated advertising and video *** major challenge lies in the fact that the target object is neither present nor localized in the input,and furthermore,available datasets only provide scenes with existing *** tackle this problem,we build an unsupervised algorithm based on object-level contexts,which explicitly models the joint probability distribution of object categories and bounding boxes using a Gaussian mixture *** on our own annotated test set demonstrate that our system outperforms existing baselines on all sub-tasks,and does so using a unified *** extensions and applications are suggested.

关键词： object recommendation bounding box prediction image composition object-level context

Spiking neural networks in intelligent control systems:a perspective

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第7期67卷 324-326页

作者： Anguo zhang Yongduan song School of Artificial Intelligence Anhui University School of Automation Chongqing University

Artificial intelligence (AI) has undergone substantial advancements in recent years, prompting an era of innovation in various domains. Among the groundbreaking contributions, neural networks stand at the forefront, powering applications ranging from computer vision and natural language processing to intelligent control systems [1, 2].

关键词：

XGCN:a library for large-scale graph neural network recommendations

在线全文

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2024年第3期18卷 247-249页

作者： Xiran song Hong HUANG Jianxun LIAN hai JIN National Engineering Research Center for Big Data Technology and System Services Computing Technology and System LabCluster and Grid Computing LabSchool of Computer Science and TechnologyHuazhong University of Science and TechnologyWuhan 430074China Microsoft Research Asia Beijing 100080China

1 Introduction Graph Neural Networks(GNNs)have gained widespread adoption in recommendation systems,and nowadays there is a pressing need to effectively manage large-scale graph data[1].When it comes to large graphs,GNNs may encounter the scalability issue stemming from their multi-layer messagepassing ***,scaling GNNs has emerged as a crucial research area in recent years,with numerous scaling strategies being proposed.

关键词： scaling gained operations

Design optimization of a hexapod vibration isolation system for electro-optical payload

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2024年第2期37卷 330-342页

作者： Jing LI Weipeng LI Xiaoyan zhang hai HUANG School of Astronautics Beihang UniversityBeijing 100191China

The electro-optical payloads on mobile platforms generally suffer undesirable vibrations generated by maneuvers and *** vibrations are in six degrees of freedom and cause line-of-sight jitters,resulting in image blurring and loss of tracking *** this paper,a Hexapod Vibration Isolation System(HVIS)is proposed and optimized to solve this *** optimization aims to centralize and minimize the natural frequencies of HVIS,for expanding the vibration isolation bandwidth and improving the vibration isolation in the higher frequency *** that the design space for HVIS is limited and interfered with the frames of the mobile platform,a non-collision algorithm is proposed and applied in the optimization to obtain the feasible optimal *** optimization result shows that the natural frequency bandwidth has been reduced by 42.9%,and the maximum natural frequency is reduced by 30.2%.The prototypes of initial and optimal designs are manufactured and *** simulated and experimental results demonstrate the validity of the optimization,and the optimal design provides a maximum of 15 dB more isolation in rotation direction than the initial design.

关键词： Electro-optical payload Vibrations Isolation system Hexapod Optimization Non-collision design

Discovery of the choristodere Hyphalosaurus baitaigouensis from the Lower Cretaceous Jiufotang Formation of western Liaoning,China and its paleobiogeographic significance

在线全文

学校读者我要写书评

暂无评论

Journal of Palaeogeography 2024年第3期13卷 547-562页

作者： Li-Jun zhang Shu-An Ji hai-Chun zhang South China Sea Culture Museum Hainan Tropical Ocean University Key Laboratory of Stratigraphy and Paleontology Ministry of Natural Resources Institute of Geology Chinese Academy of Geological Sciences State Key Laboratory of Palaeobiology and Stratigraphy Nanjing Institute of Geology and Palaeontology and Center for Excellence in Life and Paleoenvironment Chinese Academy of Sciences

The water-living reptile genus Hyphalosaurus was previously discovered in the Lower Cretaceous Yixian Formation bearing the middle Jehol Biota in western Liaoning, China. Two complete skeletons of Hyphalosaurus are reported for the first time from the Jiufotang Formation bearing the late Jehol Biota at two sites in western Liaoning. They show the typical characteristics of Hyphalosaurus baitaigouensis in terms of the contact between some cranial bones and the number of cervical and dorsal vertebrae, and therefore are assigned to H. baitaigouensis. This discovery extends the stratigraphic range of this species from the Yixian Formation to the overlying Jiufotang Formation. This paper further briefly compares the main characteristics of the two skeletons, summarizes the temporal and spatial distribution of Hyphalosaurus, and discusses the palaeogeographic evolution of western Liaoning at the middle and late Jehol Biota stages based on the aquatic habits of Hyphalosaurus and the geological background of western Liaoning. The analytical results show that the volcanic downwarped lakes, with distinct geographical divisions in western Liaoning during the deposition of the Yixian Formation, had not completely disappeared till the deposition of the Jiufotang Formation. The palaeo-lake in the Fuxin-Yixian Basin underwent the process of generation, expansion and shrinkage during the depositional period of the Yixian Formation and finally merged into a large lake. The palaeo-Dalinghe River,which connected the main basins in western Liaoning during the depositional period of the Jiufotang Formation, was formed simultaneously.

关键词： Western liaoning Jehol biota Jiufotang formation Hyphalosaurus baitaigouensis Palaeobiogeography

An optimization-oriented modeling approach using input convex neural networks and its application on optimal chiller loading

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Building Simulation 2024年第4期17卷 639-655页

作者： Shanshuo Xing Jili zhang song Mu Dalian University of Technology Dalian116024China Guangdong Airport Baiyun Information Technology Co. Ltd.Guangzhou510000China

Optimization for the multi-chiller system is an indispensable approach for the operation of highly efficient chiller *** optima obtained by model-based optimization algorithms are dependent on precise and solvable objective *** classical neural networks cannot provide convex input-output mappings despite capturing impressive nonlinear fitting capabilities,resulting in a reduction in the robustness of model-based *** this paper,we leverage the input convex neural networks(ICNN)to identify the chiller model to construct a convex mapping between control variables and the objective function,which enables the NN-based OCL as a convex optimization problem and apply it to multi-chiller optimization for optimal chiller loading(OCL).Approximation performances are evaluated through a four-model comparison based on an experimental data set,and the statistical results show that,on the premise of retaining prior convexities,the proposed model depicts excellent approximation power for the data set,especially the unseen ***,the ICNN model is applied to a typical OCL problem for a multi-chiller system and combined with three types of optimization *** with conventional and meta-heuristic methods,the numerical results suggest that the gradient-based BFGS algorithm provides better energy-saving ratios facing consecutive cooling load inputs and an impressive convergence speed.

关键词： chiller plant input convex neural network optimal load distribution convex optimization

FSCIL-EACA: Few-Shot Class-Incremental Learning Network Based on Embedding Augmentation and Classifier Adaptation for Image Classification

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2024年第1期33卷 139-152页

作者： Ruru zhang haihong E Meina song School of Computer Science Beijing University of Posts and Telecommunications Education Department Information Network Engineering Research Center Beijing University of Posts and Telecommunications

The ability to learn incrementally is critical to the long-term operation of AI systems. Benefiting from the power of few-shot class-incremental learning(FSCIL), deep learning models can continuously recognize new classes with only a few samples. The difficulty is that limited instances of new classes will lead to overfitting and exacerbate the catastrophic forgetting of the old classes. Most previous works alleviate the above problems by imposing strong constraints on the model structure or parameters, but ignoring embedding network transferability and classifier adaptation(CA), failing to guarantee the efficient utilization of visual features and establishing relationships between old and new classes. In this paper, we propose a simple and novel approach from two perspectives: embedding bias and classifier bias. The method learns an embedding augmented(EA) network with cross-class transfer and class-specific discriminative abilities based on self-supervised learning and modulated attention to alleviate embedding bias. Based on the adaptive incremental classifier learning scheme to realize incremental learning capability,guiding the adaptive update of prototypes and feature embeddings to alleviate classifier bias. We conduct extensive experiments on two popular natural image datasets and two medical datasets. The experiments show that our method is significantly better than the baseline and achieves state-of-the-art results.

关键词： Training Adaptation models Visualization Image recognition Prototypes Self-supervised learning Power capacitors

Learning topological defects formation with neural networks in a quantum phase transition

在线全文

学校读者我要写书评

暂无评论

Communications in Theoretical Physics 2024年第5期76卷 68-76页

作者： Han-Qing Shi hai-Qing zhang Center for Gravitational Physics Department of Space ScienceBeihang UniversityBeijing 100191China Peng Huanwu Collaborative Center for Research and Education Beihang UniversityBeijing 100191China

Neural networks possess formidable representational power,rendering them invaluable in solving complex quantum many-body *** they excel at analyzing static solutions,nonequilibrium processes,including critical dynamics during a quantum phase transition,pose a greater challenge for neural *** address this,we utilize neural networks and machine learning algorithms to investigate time evolutions,universal statistics,and correlations of topological defects in a one-dimensional transverse-field quantum Ising ***,our analysis involves computing the energy of the system during a quantum phase transition following a linear quench of the transverse magnetic field *** excitation energies satisfy a power-law relation to the quench rate,indicating a proportional relationship between the excitation energy and the kink ***,we establish a universal power-law relationship between the first three cumulants of the kink numbers and the quench rate,indicating a binomial distribution of the ***,the normalized kink-kink correlations are also investigated and it is found that the numerical values are consistent with the analytic formula.

关键词： neural networks machine learning transverse-field quantum Ising model kibble-zurek mechanism