检索结果-内蒙古大学图书馆

Modified Zhang neural algorithm for time-varying overdetermined linear models and applications

APPLIED MATHEMATICAL MODELLING 2023年 114卷 44-60页

作者： Wu, Dongqing Zhang, Yunong Zhongkai Univ Agr & Engn Sch Computat Sci Guangzhou 510220 Guangdong Peoples R China Sun Yat Sen Univ Sch Comp Sci & Engn Guangzhou 510006 Guangdong Peoples R China

The challenges of the inefficiency of the existing static least-squares method and the in-sufficient precision of the existing neural algorithms need to be overcome. With the aid of the ten-instant Zhang time discretization formula, a novel ten-instant discrete-time neural network algorithm is proposed and generalized to the problem-solving of overdetermined system of time-varying linear equations. Theoretical analyses illustrate that the maximal steady-state residual error of the proposed algorithm has an order-6 pattern. Comparative experiment results further substantiate the superiority of the proposed algorithm. More-over, applications of mobile localization based on the angle of arrival technique and the time difference of arrival technique are provided to display the applicability of the pro-posed algorithm.(c) 2022 Elsevier Inc. All rights reserved.

关键词： Least -squares method Zhang neural network neural algorithm Mobile localization

来源：评论

学校读者我要写书评

暂无评论

A Noise-Suppressing neural algorithm for Solving the Time-Varying System of Linear Equations: A Control-Based Approach

引用

IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS 2019年第1期15卷 236-246页

作者： Jin, Long Li, Shuai Hu, Bin Liu, Mei Yu, Jiguo Lanzhou Univ Sch Informat Sci & Engn Lanzhou 730000 Gansu Peoples R China Hong Kong Polytech Univ Dept Comp Hung Hom Hong Kong Peoples R China Qufu Normal Univ Sch Informat Sci & Engn Rizhao 276800 Peoples R China

It has been found that there exists an essential similarity between solving equations and controlling dynamic systems: Both errors are expected to decrease to zero (or an acceptably tiny value) as soon as possible. By exploiting such a similarity, researchers have presented and investigated continuous-time recurrent neural network models for solving time-varying problems. To be compatible with digital computers, it is desirable to develop discrete-time neural algorithms from the control perspective for performance improvement. In this paper, a discrete-time zeroing neural algorithm is proposed for the solving system of linear equations with the aid of control techniques. To lay a basis for theoretical analyses, the proposed zeroing neural algorithm with nonlinearity is converted into a second-order linear system plus a residual term, and then, analyzed using the control theory. Theoretical results and numerical experiments are provided, which illustrate that the proposed neural algorithm possesses an improved performance compared to the existing solutions.

关键词： Control techniques neural algorithm noise reduction residual error system of linear equations

来源：评论

学校读者我要写书评

暂无评论

Digraph visualization using a neural algorithm with a heuristic activation scheme

引用

IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS 1998年第4期28卷 562-572页

作者： Kusnadi Beebe, C Carothers, JD Univ Arizona Dept Elect & Comp Engn Tucson AZ 85721 USA

This paper presents an activation scheme for use with Hopfield neural network algorithms that guarantees a valid solution for a particular category of problems. The technique monitors the appropriate neurons and heuristically controls their activation function. As a result it has been possible to eliminate several constraint terms from the energy function that normally would have been required to drive the network toward a valid solution. This saves time and eliminates the need for empirically determining a larger number of constants. This technique has been applied to the combinatorial optimization problem called hierarchical digraph visualization that arises in many application areas where it is necessary to visually realize the relationship between entities in complex systems. Results are presented that compare this new approach with a more traditional neural network approach as well as heuristic approaches. Performance improvement in terms of the solution quality as well as execution time relative to both alternative techniques was achieved.

关键词： combinatorial optimization hierarchical digraph visualization Hopfield network neural algorithm

来源：评论

学校读者我要写书评

暂无评论

A neural algorithm for computing bipartite matchings

引用

PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA 2024年第37期121卷 e2321032121-e2321032121页

作者： Dasgupta, Sanjoy Meirovitch, Yaron Zheng, Xingyu Bush, Inle Lichtman, Jeff W. Navlakha, Saket Univ Calif San Diego Comp Sci & Engn Dept La Jolla CA 92037 USA Harvard Univ Dept Mol & Cell Biol Cambridge MA 02138 USA Harvard Univ Ctr Brain Sci Cambridge MA 02138 USA Simons Ctr Quantitat Biol Cold Spring Harbor Lab Cold Spring Harbor NY 11724 USA

Finding optimal bipartite matchings-e.g., matching medical students to hospitals for residency, items to buyers in an auction, or papers to reviewers for peer review-is a fundamental combinatorial optimization problem. We found a distributed algorithm for computing matchings by studying the development of the neuromuscular circuit. The neuromuscular circuit can be viewed as a bipartite graph formed between motor neurons and muscle fibers. In newborn animals, neurons and fibers are densely connected, but after development, each fiber is typically matched (i.e., connected) to exactly one neuron. We cast this synaptic pruning process as a distributed matching (or assignment) algorithm, where motor neurons "compete" with each other to "win" muscle fibers. We show that this algorithm is simple to implement, theoretically sound, and effective in practice when evaluated on real-world bipartite matching problems. Thus, insights from the development of neural circuits can inform the design of algorithms for fundamental computational problems.

关键词： neural algorithm bipartite matching neuromuscular circuit circuit development neural-inspired computing

来源：评论

学校读者我要写书评

暂无评论

A neural algorithm TO SOLVE THE HAMILTONIAN CYCLE PROBLEM

A NEURAL ALGORITHM TO SOLVE THE HAMILTONIAN CYCLE PROBLEM

引用

INTERNATIONAL JOINT CONF ON neural NETWORKS ( IJCNM-90 )

作者： MEHTA, S FULOP, L Electr Comput & Syst Eng Dept Rensselaer Polytechnic Inst Troy NY USA

A network of analog neurons to solve the Hamiltonian cycle problem (HCP) is described. This neural net is a modification of the network proposed by Hopfield to solve the traveling salesman problem (TSP). A result on the convergence of quasi-stationary flow and a bound for the strength of an inhibitory self-connection are presented. Results of successful experiments with graphs of up to 500 nodes are reported. The result of an experiment with the 318-city TSP is also reported. Contrary to intuition, the performance improves with the size of the graphs. The 20-node graphs fail to give consistent results for 10 % connectivity, while 400- and 500-node graphs were solved successfully

关键词： neural nets Hopfield network neural algorithm Hamiltonian cycle problem traveling salesman problem inhibitory self-connection neural nets operations research

来源：评论

学校读者我要写书评

暂无评论

Modification of neural Network Tsang-Wang in algorithm for CAD of Complex Systems with Higher Degree of Dependability 1

引用

11th International Conference on Dependability and Complex Systems (DepCoS-RELCOMEX)

作者： Drabowski, Mieczyslaw Cracow Univ Technol Fac Elect & Comp Engn PL-30084 Krakow Poland

ISBN: (数字)9783319396392

ISBN: (纸本)9783319396392;9783319396385

The paper includes a proposal of a new algorithm for Computer Aided Design (CAD) of complex system with higher degree of dependability. Optimal scheduling of tasks and optimal resources partition are fundamental problems in this algorithm. Presented the CAD algorithm, based on neural networks, may have a practical application in developing tools for rapid prototyping of such systems.

关键词： Complex system Scheduling Partition Allocation Dependable Optimization neural algorithm Tsang-Wang networks CAD tools

来源：评论

学校读者我要写书评

暂无评论

A Review: neural-Inspired Photonic Functional Systems for Dynamic RF Signal Processing

引用

JOURNAL OF LIGHTWAVE TECHNOLOGY 2020年第19期38卷 5318-5326页

作者： Fok, Mable P. Univ Georgia Coll Engn Lightwave & Microwave Photon Lab Athens GA 30602 USA

Drawing inspiration from our nature, photonic systems that use light to imitate neural algorithms and behaviors of nature could be effective to solve complex problems in human's civilized society that have been challenging for conventional electronic to tackle. Since neural algorithms are natural designs that have undergone hundreds of million years of evolution and govern the survival of the organism, therefore, those neural algorithms are highly effective for the designated tasks. The goal of this article is to review the recent demonstrations of the photonic implementations of small scale functional neural algorithms for dynamic RF signal processing applications. In this article, two small-scale neural algorithms are reviewed - (i) spike timing dependent plasticity, an algorithm that governs howneural network are connecting together and how learning/adaptation can be achieved in animals, and (ii) jamming avoidance response in Eigenmannia, an algorithm in a gene of electric fish that mitigates frequency jamming between neighboring electric fish. The photonic circuits that are inspired by the two neural algorithms are also presented and the real-life applications of the neural algorithms in human society will be discussed.

关键词： neural-inspired photonics biomimetic photonics photonic neuron bio-inspired signal processing neural algorithm

来源：评论

学校读者我要写书评

暂无评论

Self-reconstruction of 3D mesh arrays with 11/2-track switches by digital neural circuits

引用

IEICE TRANSACTIONS ON ELECTRONICS 1999年第9期E82C卷 1678-1686页

作者： Takanami, I Nakamura, S Horita, T Iwate Univ Fac Engn Morioka Iwate 0208551 Japan ICS Co Morioka Iwate 0208544 Japan

Using Hopfield-type neural network model, we present an algorithm for reconstructing 3D mesh processor arrays using single-track switches where spare processors are laid on the six surfaces of a 3D array and show its effectiveness in terms of reconstruction rate and computing time by computer simulation. Next, we show how the algorithm can be realized by a digital neural circuit. It consists of subcircuits for finding candidate compensation paths, deciding whether the neural system reaches a stable state and at the time the system energy is minimum, and subcircuits for neurons. The subcircuit for each neuron including the other subcircuits can only be made with 16 gates and two flip-flops. Since the state transitions are done in parallel, the circuit will be able to find a set of compensation paths for a fault pattern very quickly within a time less than l mu s. Furthermore, the hardware implementation of the algorithm leads to making a self-reconfigurable system without the aid of a host computer.

关键词： fault tolerance three-dimensional mesh array self-reconstruction neural algorithm neural circuit

来源：评论

学校读者我要写书评

暂无评论

The use of artificial neural networks in analysing the nutritional ecology of Chrysomya megacephala (F.) (Diptera: Calliphoridae), compared with a statistical model

引用

AUSTRALIAN JOURNAL OF ENTOMOLOGY 2010年第3期49卷 201-212页

作者： Bianconi, Andre Von Zuben, Claudio J. de Souza Serapiao, Adriane B. Govone, Jose S. Sao Paulo State Univ UNESP Inst Biociencias Dept Zool BR-13506900 Rio Claro SP Brazil Dept Estat Matemat Aplicada & Comp Rio Claro SP Brazil

Artificial neural networks (ANNs) have been widely applied to the resolution of complex biological problems. An important feature of neural models is that their implementation is not precluded by the theoretical distribution shape of the data used. Frequently, the performance of ANNs over linear or non-linear regression-based statistical methods is deemed to be significantly superior if suitable sample sizes are provided, especially in multidimensional and non-linear processes. The current work was aimed at utilising three well-known neural network methods in order to evaluate whether these models would be able to provide more accurate outcomes in relation to a conventional regression method in pupal weight predictions of Chrysomya megacephala, a species of blowfly (Diptera: Calliphoridae), using larval density (i.e. the initial number of larvae), amount of available food and pupal size as input data. It was possible to notice that the neural networks yielded more accurate performances in comparison with the statistical model (multiple regression). Assessing the three types of networks utilised (Multi-layer Perceptron, Radial Basis Function and Generalised Regression neural Network), no considerable differences between these models were detected. The superiority of these neural models over a classical statistical method represents an important fact, because more accurate models may clarify several intricate aspects concerning the nutritional ecology of blowflies.

关键词： blowfly larval density mass rearing neural algorithm pupal weight

来源：评论

学校读者我要写书评

暂无评论

Image Synthesis using Convolutional neural Network 2

Image Synthesis using Convolutional Neural Network

引用

2nd IEEE International Conference on Recent Trends in Electronics, Information and Communication Technology (RTEICT)

作者： Bhat, Ganesh Dharwadkar, Shrikant Reddy, N. V. Subba Shivaprasad, G. Manipal Univ MIT Dept CSE Manipal Karnataka India

ISBN: (纸本)9781509037049

The image processing task for conversion of image style is a tedious task. The previous approaches carried out in this area lacked in extracting higher level of abstractions from the input image, hence failed to generate image content from the acquired knowledge. Hence to achieve higher level knowledge from the object to be recognized, we make use of convolutional neural networks which aid in achieving higher quality images. The proposed work uses a neural algorithm that produces an image with a new style using the knowledge acquired from input image. Our work provides a new approach for deep representations of images using neural Networks, hence demonstrating the generation of high level image synthesis and manipulation.

关键词： Image synthesis Image processing neural algorithm Convolutional neural network Knowledge extraction

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：