共查询到19条相似文献,搜索用时 171 毫秒
1.
Zhi-HuaZhou 《计算机科学技术学报》2004,19(C00):19-19
传统的神经网络是一种黑箱模型,其学习到的知识蕴涵在大量连接权中,这严重阻碍了神经网络技术在对可理解性要求较强的领域的应用。由于从神经网络中抽取出易于理解的符号规则有助于解决该问题,因此,神经网络规则抽取在上世纪90年代成为神经计算界的一个研究热点。 相似文献
2.
3.
4.
讨论了采用BP(Back—Propagation)网络来识别给定的心电图的有关技术。这一课题涉及多个知识领域,神经计算、模式识别、数值计算及心电图学等。构成系统的主要思想是:将给定的心电采样数据经过K—L展开后,提取出有关给定数据的特征向量,并将此特征向量作为神经网络的输入,神经网络经过一系列的学习,将这些特征信息转换成相应的知识规则,存放在由网络形成的知识库中。在识别给定信号时,首先也是进行给定数据特征抽取,然后将所得的特征向量输入神经网络,神经网络运用知识库中存放的规则来识别给出的数据,并输出相关的信息。 相似文献
5.
神经网络规则抽取是神经网络领域的一个重要方向,但是对抽取的规则评估算法却很少.针对这一问题,提出了神经网络抽取规则评估方法.首先证明所有的规则形式都可以统一为区间的形式,然后在区间算法的基础上提出规则评估方法.评估的标准有四个:覆盖性、准确性、矛盾性,以及冗余性.由于规则的矛盾性和冗余性是规则之间的问题,所以该文仅仅研究规则的覆盖性和准确性,提出了覆盖性判断定理,并提出了覆盖性、准确性判断算法.实例证实了该算法的有效性. 相似文献
6.
面向神经计算的并行计算机体系结构是神经网络研究中的一项重要工作。本文在对大量的神经计算进行需求分析的基础上,讨论了以高性能的微处理器作为计算单元,进行面向神经计算的并行计算机体系结构设计,并且介绍了它原型实现的结构、参数和性能 相似文献
7.
8.
9.
10.
赵艳 《计算机与信息技术》2009,(5)
神经计算科学是从信息科学的角度,用计算的方法研究神经网络如何模仿和延伸人脑活动的机理及实现类脑智能信息系统的问题。量子神经计算是量子计算与神经计算相结合的产物。文中主要阐述了神经计算的研究现状,在其基础上对量子神经计算的概念及模型进行了介绍,综述了国内外的研究动态与发展趋势。 相似文献
11.
Henry Tirri 《New Generation Computing》1991,10(1):55-71
The relation of subsymbolic (neural computing) and symbolic computing has been a topic of intense discussion. We address some
of the drawbacks of current expert system technology and study the possibility of using neural computing principles to improve
their competence. In this paper we focus on the problem of using neural networks to implement expert system rule conditions.
Our approach allows symbolic inference engines to make direct use of complex sensory input via so called detector predicates.
We also discuss the use of self organizing Kohonen networks as a means to determine those attributes (properties) of data
that reflect meaningful statistical relationships in the expert system input space. This mechanism can be used to address
the defficult problem of conceptual clustering of information. The concepts introduced are illustrated by two application
examples: an automatic inspection system for circuit packs and an expert system for respiratory and anesthesia monitoring.
The adopted approach differs from the earlier research on the use of neural networks as expert systems, where the only method
to obtain knowledge is learning from training data. In our approach the synergy of rules and detector predicates combines
the advantages of both worlds: it maintains the clarity of the rule-based knowledge representation at the higher reasoning
levels without sacrificing the power of noise-tolerant pattern association offered by neural computing methods.
This research is supported by Technology Development Center (TEKES) in Software Technology Programme (FINSOFT). Part of this
work was done while the author was visiting AT & T Bell Laboratories. 相似文献
12.
Wangli HAO Ian Max ANDOLINA Wei WANG Zhaoxiang ZHANG 《Frontiers of Computer Science》2021,15(1):151304-15
Visual information is highly advantageous for the evolutionary success of almost all animals.This information is likewise critical for many computing tasks,and visual computing has achieved tremendous successes in numerous applications over the last 60 years or so.In that time,the development of visual computing has moved forwards with inspiration from biological mechanisms many times.In particular,deep neural networks were inspired by the hierarchical processing mechanisms that exist in the visual cortex of primate brains(including ours),and have achieved huge breakthroughs in many domainspecific visual tasks.In order to better understand biologically inspired visual computing,we will present a survey of the current work,and hope to offer some new avenues for rethinking visual computing and designing novel neural network architectures. 相似文献
13.
人脸表情识别已成为人工智能领域的重要研究课题,但传统的卷积神经网络需要庞大的计算资源使得其应用受限,而二值化卷积神经网络可通过快速与或运算代替原本的浮点乘法运算,大大降低了算法对计算资源的需求。论文提出了一种基于数据增强和二值化卷积神经网络的人脸表情识别算法,通过均值估计,在FER2013数据集上达到了66.15%的识别率,超越了部分基于浮点乘积运算的卷积网络,为表情识别算法移植到小型设备中提供了可能。 相似文献
14.
为降低集成特征选择方法的计算复杂性,提出了一种基于粗糙集约简的神经网络集成分类方法。该方法首先通过结合遗传算法求约简和重采样技术的动态约简技术,获得稳定的、泛化能力较强的属性约简集;然后,基于不同约简设计BP网络作为待集成的基分类器,并依据选择性集成思想,通过一定的搜索策略,找到具有最佳泛化性能的集成网络;最后通过多数投票法实现神经网络集成分类。该方法在某地区Landsat 7波段遥感图像的分类实验中得到了验证,由于通过粗糙集约简,过滤掉了大量分类性能欠佳的特征子集,和传统的集成特征选择方法相比,该方法时 相似文献
15.
为降低集成特征选择方法的计算复杂性,提出了一种基于粗糙集约简的神经网络集成分类方法。该方法首先通过结合遗传算法求约简和重采样技术的动态约简技术,获得稳定的、泛化能力较强的属性约简集;然后,基于不同约简设计BP网络作为待集成的基分类器,并依据选择性集成思想,通过一定的搜索策略,找到具有最佳泛化性能的集成网络;最后通过多数投票法实现神经网络集成分类。该方法在某地区Landsat 7波段遥感图像的分类实验中得到了验证,由于通过粗糙集约简,过滤掉了大量分类性能欠佳的特征子集,和传统的集成特征选择方法相比,该方法时间开销少,计算复杂性低,具有满意的分类性能。 相似文献
16.
脉冲神经网络属于第三代人工神经网络,它是更具有生物可解释性的神经网络模型。随着人们对脉冲神经网络不断深入地研究,不仅神经元空间结构更为复杂,而且神经网络结构规模也随之增大。以串行计算的方式,难以在个人计算机上实现脉冲神经网络的模拟仿真。为此,设计了一个多核并行的脉冲神经网络模拟器,对神经元进行编码与映射,自定义路由表解决了多核间的网络通信,以时间驱动为策略,实现核与核间的动态同步,在模拟器上进行脉冲神经网络的并行计算。以Izhikevich脉冲神经元为模型,在模拟环境下进行仿真实验,结果表明多核并行计算相比传统的串行计算在效率方面约有两倍的提升,可为类似的脉冲神经网络的模拟并行化设计提供参考。 相似文献
17.
K.M. Saridakis 《Advanced Engineering Informatics》2008,22(2):202-221
The present paper surveys the application of soft computing (SC) techniques in engineering design. Within this context, fuzzy logic (FL), genetic algorithms (GA) and artificial neural networks (ANN), as well as their fusion are reviewed in order to examine the capability of soft computing methods and techniques to effectively address various hard-to-solve design tasks and issues. Both these tasks and issues are studied in the first part of the paper accompanied by references to some results extracted from a survey performed for in some industrial enterprises. The second part of the paper makes an extensive review of the literature regarding the application of soft computing (SC) techniques in engineering design. Although this review cannot be collectively exhaustive, it may be considered as a valuable guide for researchers who are interested in the domain of engineering design and wish to explore the opportunities offered by fuzzy logic, artificial neural networks and genetic algorithms for further improvement of both the design outcome and the design process itself. An arithmetic method is used in order to evaluate the review results, to locate the research areas where SC has already given considerable results and to reveal new research opportunities. 相似文献
18.
We introduce a framework for simulating signal propagation in geometric networks (networks that can be mapped to geometric graphs in some space) and developing algorithms that estimate (i.e., map) the state and functional topology of complex dynamic geometric networks. Within the framework, we define the key features typically present in such networks and of particular relevance to biological cellular neural networks: dynamics, signaling, observation, and control. The framework is particularly well suited for estimating functional connectivity in cellular neural networks from experimentally observable data and has been implemented using graphics processing unit high-performance computing. Computationally, the framework can simulate cellular network signaling close to or faster than real time. We further propose a standard test set of networks to measure performance and compare different mapping algorithms. 相似文献
19.
Multilayer perceptrons (MLPs) with weight values restricted to powers of two or sums of powers of two are introduced. In a digital implementation, these neural networks do not need multipliers but only shift registers when computing in forward mode, thus saving chip area and computation time. A learning procedure, based on backpropagation, is presented for such neural networks. This learning procedure requires full real arithmetic and therefore must be performed offline. Some test cases are presented, concerning MLPs with hidden layers of different sizes, on pattern recognition problems. Such tests demonstrate the validity and the generalization capability of the method and give some insight into the behavior of the learning algorithm 相似文献