检索结果-内蒙古大学图书馆

Efficient mapping of backpropagation algorithm onto a network of workstations

IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS 1998年第6期28卷 841-848页

作者： Sudhakar, V Murthy, CSR Indian Inst Technol Dept Comp Sci & Engn Madras 600036 Tamil Nadu India

In this paper, we present an efficient technique for mapping a backpropagation (BP) learning algorithm for multilayered neural networks onto a network of workstations (NOW's). We adopt a vertical partitioning scheme, where each layer in the neural network is divided into p disjoint partitions, and map each partition onto an independent workstation in a network of p workstations. We present a fully distributed version of the BP algorithm and also its speedup analysis. We compare the performance of our algorithm with a recent work involving the vertical partitioning approach for mapping the BP algorithm onto a distributed memory multiprocessor. Our results on SUN 3/50 NOW's show that we are able to achieve better speedups by using only two communication sets and also by avoiding some redundancy in the weights computation for one training cycle of the algorithm.

关键词： backpropagation algorithm distributed memory multiprocessors multilayered neural networks network of workstations network partitioning pattern partitioning performance analysis

来源：评论

学校读者我要写书评

暂无评论

Parameter identification of proton exchange membrane fuel cell via Levenberg-Marquardt backpropagation algorithm

引用

INTERNATIONAL JOURNAL OF HYDROGEN ENERGY 2021年第44期46卷 22998-23012页

作者： Yang, Bo Zeng, Chunyuan Wang, Long Guo, Yinyuan Chen, Guanghua Guo, Zhengxun Chen, Yijun Li, Danyang Cao, Pulin Shu, Hongchun Yu, Tao Zhu, Jiawei Kunming Univ Sci & Technol Fac Elect Power Engn Kunming 650500 Yunnan Peoples R China Zhengzhou Univ Aeronaut Zhengzhou 450046 Peoples R China Xuchang KETOP Testing Res Inst Co Xuchang Peoples R China South China Univ Technol Coll Elect Power Guangzhou 510640 Peoples R China Changan Univ Sch Informat Engn Xian 710064 Peoples R China

It is essential to develop an accurate model of proton exchange membrane fuel cell (PEMFC) for a reliable operation and analysis, in which unknown parameters usually need to be determined. The inherent nonlinear, strong coupling, and diversification of PEMFC model seriously hinder traditional methods to identify the parameters. For the sake of overcoming these thorny obstacles, Levenberg-Marquardt backpropagation (LMBP) algorithm based on artificial neural networks (ANNs) is proposed for PEMFC parameter identification. Furthermore, the performance of LMBP is thoroughly evaluated and compared with four typical meta-heuristic algorithms under three cases. Simulation results indicate that LMBP performs a higher accuracy and faster speed for parameter identification. In particular, accuracy and convergence speed can achieve as much as 99.8% and 95.9% growth via LMBP, respectively. (c) 2021 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

关键词： Proton exchange membrane fuel cell Parameter identification Levenberg-marquardt backpropagation algorithm Artificial neural networks Meta-heuristic algorithm

来源：评论

学校读者我要写书评

暂无评论

A SCALABLE PARALLEL FORMULATION OF THE backpropagation algorithm FOR HYPERCUBES AND RELATED ARCHITECTURES

引用

IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS 1994年第10期5卷 1073-1090页

作者： KUMAR, V SHEKHAR, S AMIN, MB Department of Computer Science University of Minnesota Minneapolis MN USA

In this paper, we present a new technique for mapping the backpropagation algorithm on hypercubes and related architectures. A key component of this technique is a network partitioning scheme called checkerboarding, Checkerboarding allows ns to replace the ail-to-all broadcast operation performed by the commonly used vertical network partitioning scheme, with operations that are much faster on the hypercubes and related architectures. Checkerboarding can be combined with the pattern partitioning technique to form a hybrid scheme that performs better than either one of these schemes. Theoretical analysis and experimental results on nCUBE(R) and CM5(R) show that our scheme performs better than the other schemes, for both uniform and nonuniform networks.

关键词： backpropagation algorithm HYPERCUBE SCALABILITY NEURAL NETWORKS NETWORK PARTITIONING PATTERN PARTITIONING

来源：评论

学校读者我要写书评

暂无评论

MODIFIED backpropagation algorithm FOR FAST LEARNING IN NEURAL NETWORKS

引用

ELECTRONICS LETTERS 1990年第19期26卷 1564-1566页

作者： REYNERI, LM FILIPPI, E Dipartimento di Elettronica Politecnico di Torino Torino Italy

A fast learning rule for artificial neural systems which is based on modifications to a backpropagation algorithm is described. The rule minimises the error function along the direction of the gradient and backpropaga... 详细信息

关键词： neural nets artificial neural systems Artificial intelligence Adaptive system theory adaptive systems fast learning rule backpropagation algorithm virtual machines constant error energy approach

来源：评论

学校读者我要写书评

暂无评论

Controllable pitch propeller control with fast backpropagation algorithm

引用

MARINE TECHNOLOGY AND SNAME NEWS 2007年第3期44卷 180-184页

作者： Alarcin, Fuat Ekinci, Serkan Gulez, Kayhan Yildiz Tech Univ Dept Naval Architecture & Marine Engn Istanbul Turkey Yildiz Tech Univ Dept Elect Engn Istanbul Turkey

In this paper, controllable pitch propeller (CPP) has been used to improve the maneuvering at a fixed shaft speed. The proportional-derivative (PD) control method has been applied to obtain high efficiency by changing the pitch angle. The controller results have been evaluated and learned by a fast backpropagation algorithm (FBA). The neural network results show that learning of pitch behavior has been verified, and the FBA performance of CPP controller is better than that of the PD controller.

关键词： neural network shaft speed control system backpropagation algorithm Engineering controllable pitch propeller propulsion system machine learning controller controllable pitch propeller control Information Management and Systems Neural networks

来源：评论

学校读者我要写书评

暂无评论

Drowsy driving detection using neural network with backpropagation algorithm implemented by FPGA

引用

CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE 2020年第18期32卷

作者： Choi, Hyun-Sik Chosun Univ Coll IT Convergence Engn Dept Elect Engn Gwangju 61452 South Korea

Neural networks have recently attracted much attention due to the development of artificial intelligence or deep learning technology. These can be implemented by applying the current hardware technology such as a central processing unit and a graphics processing unit. In this case, the applications are limited because considerable power and large volume are used. To overcome these shortcomings, hardware development for artificial intelligence is accelerated, and this technology is called the neuromorphic system, which is especially suitable for low-power and small-area applications such as wearable devices. In this study, the neuromorphic system is implemented using the field-programmable gate array (FPGA), and it is applied to wearable systems. This system is especially developed for a module that measures the drowsiness of a user based on biosignals such as electrocardiogram (ECG) and electromyography (EMG). The measured biosignals are fed to the neuromorphic system for supervised learning using the backpropagation algorithm. Therefore, it is possible to make the drowsiness driving assessment specific to each user, and the error on the user's condition can be minimized. In addition, by integrating artificial intelligence including learning algorithm and biosensor circuits, it is possible to minimize disturbance to the driver or user through miniaturization and low power consumption.

关键词： backpropagation algorithm drowsiness driving assessment electrocardiogram (ECG) electromyography (EMG) field-programmable gate array (FPGA) neural networks neuromorphic system wearable devices

来源：评论

学校读者我要写书评

暂无评论

A Study on Lyapunov Function backpropagation algorithm for Suitability as Neuro-Adaptive Inverse Controller 21

A Study on Lyapunov Function Backpropagation Algorithm for S...

引用

21st IEEE International Multi-Topic Conference (INMIC)

作者： Aftab, Muhammad Saleheen Aftab, Fakhra United Arab Emirates Univ Dept Elect Engn Al Ain U Arab Emirates NED Univ Engn & Technol Comp & Informat Syst Engn Dept Karachi Pakistan

ISBN: (纸本)9781538675366

Artificial neural networks are commonly known as Universal Approximators;a property immensely useful in system identification and control applications. Traditionally, neural networks are trained with gradient-descent backpropagation algorithms. However, these algorithms are computationally burdened and slow due to the calculation of error derivatives. As a result, the research focus has shifted to develop gradient-free neural algorithms. One famous approach is to incorporate Lyapunov Functions in network parameter optimization. In this paper, we briefly discuss and analyze one such recently developed algorithm from the point-of-view of its applicability in adaptive control paradigm. It has been found that with a few proposed modifications, this algorithm can work excellently as neuroadaptive inverse controller.

关键词： artificial neural networks backpropagation algorithm Lyapunov stability theory adaptive inverse controller system identification

来源：评论

学校读者我要写书评

暂无评论

Strategies for Determining Effective Step Size of the backpropagation algorithm for On-Line Learning 7

Strategies for Determining Effective Step Size of the Backpr...

引用

7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)

作者： Kaneda, Yuya Zhao, Qiangfu Liu, Yong Pei, Yan Univ Aizu Dept Comp & Informat Syst Aizu Wakamatsu Fukushima Japan

ISBN: (纸本)9781467393607

In this paper, we investigate proper strategies for determining the step size of the backpropagation (BP) algorithm for on-line learning. It is known that for off-line learning, the step size can be determined adaptively during learning. For on-line learning, since the same data may never appear again, we cannot use the same strategy proposed for off-line learning. If we do not update the neural network with a proper step size for on-line learning, the performance of the network may not be improved steadily. Here, we investigate four strategies for updating the step size. They are 1) constant, 2) random, 3) linearly decreasing, and 4) inversely proportional, respectively. The first strategy uses a constant step size during learning, the second strategy uses a random step size, the third strategy decreases the step size linearly, and the fourth strategy updates the step size inversely proportional to time. Experimental results show that, the third and the fourth strategies are more effective. In addition, compared with the third strategy, the fourth one is more stable, and usually can improve the performance steadily.

关键词： Decision Boundary Making Multilayer Perceptron backpropagation algorithm On-Line Learning

来源：评论

学校读者我要写书评

暂无评论

An evaluation of the bihyperbolic function in the optimization of the backpropagation algorithm

引用

INTERNATIONAL TRANSACTIONS IN OPERATIONAL RESEARCH 2014年第5期21卷 835-854页

作者： Miguez, Geraldo Xavier, Adilson Elias Maculan, Nelson Univ Fed Rio de Janeiro PESC COPPE Rio De Janeiro Brazil

The backpropagation algorithm is one of the most used tools for training artificial neural networks. However, this tool may be very slow in some practical applications. Many techniques have been discussed to speed up the performance of this algorithm and allow its use in an even broader range of applications. Although the backpropagation algorithm has been used for decades, we present here a set of computational results that suggest that by replacing bihyperbolic functions the backpropagation algorithm performs better than the traditional sigmoid functions. To the best of our knowledge, this finding was never previously published in the open literature. The efficiency and discrimination capacity of the proposed methodology are shown through a set of computational experiments, and compared with the traditional problems of the literature.

关键词： optimization backpropagation algorithm machine learning neural networks activation function pattern recognition bihyperbolic function

来源：评论

学校读者我要写书评

暂无评论

A new probabilistic neural network model based on backpropagation algorithm

引用

JOURNAL OF INTELLIGENT & FUZZY SYSTEMS 2017年第1期32卷 215-227页

作者： Sun, Qian Wu, Chong Li, Yong-li Harbin Inst Technol Sch Management Harbin Peoples R China Heilongjiang Inst Sci & Technol Sch Econ Harbin Peoples R China

Since the traditional probabilistic neural network (PNN) cannot systematically solve the difficulty of estimating probability function and the high space complexity, this paper introduces backpropagation (BP) algorithm into the classical PNN. By designing appropriate error function and BP algorithm based on the steepest descent, an improved BP-PNN is presented, with its algorithm and effectiveness deduced. Three synthetic datasets and ten benchmark problems have been tested, compared with Probabilistic Neural Networks (PNN), Multi-Layered Perceptron (MLP) and Support Vector Machine (SVM). The results prove that (1) the accuracy of classification of BP-PNN is much higher than PNN, and it has a significant advantage compared with MLP and SVM;(2) BP-PNN has strong capacity to identify the importance of input indicators;(3) BP-PNN is a new pattern classification method to estimate the probabilistic function, reduce the space complexity and identify the importance of the indicators.

关键词： Probabilistic neural network backpropagation algorithm classification decision analysis

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：