摘要
基于云计算技术的神经网络方法研究为大规模数据的分析处理提供了廉价的、高效的解决方案,交叉覆盖算法可以较好地解决多层前向网络分类器的设计问题,弥补BP神经网络的不足.结合MP神经元的几何概念,基于交叉覆盖算法设计神经网络,以Iris数据集为例,基于云计算环境实现了数据分类,为多层前向神经网络在云环境下进行数据分析提供了一种有意义的研究和实践.
The research of neural network based on cloud computing provides a cheap and efficient solution for the analysis and processing of large-scale data.The alternative covering algorithm can better solve the design problem of multi-layer neural network as a classifier,and make up for the shortage of BP neural network.The neural network was designed based on the geometrical representation of MP neural model and the alternative covering algorithm,and realized the classification of Iris data sets in cloud computing.This provides a meaningful research and practice for the data analysis based on the alternative covering algorithm of multi-layer neural networks in cloud computing.
作者
张涛
ZHANG Tao(Department of Information Engineering,Tongling Polytechnic,Tongling,Anhui 244061,China)
出处
《宜宾学院学报》
2019年第6期47-51,共5页
Journal of Yibin University
基金
安徽省高校自然科学重点研究项目“基于云计算的智慧校园支撑服务平台研究”(KJ2018A0748)
安徽省高校自然科学重点研究项目“移动云计算环境下的高校主数据管理研究”(KJ2018A0751)
安徽省高校优秀青年人才支持计划项目(gxyq2019218)
关键词
多层神经网络
云计算
交叉覆盖
数据分类
multi-layer neural networks
cloud computing
alternative covering algorithm
data classification