| 高属性维稀疏数据聚类回归逻辑 神经网络模型及学习算法 |
| |
| 引用本文： | 周永权,焦李成.高属性维稀疏数据聚类回归逻辑 神经网络模型及学习算法[J].电子学报,2004,32(8):1342-1345. |
| |
| 作者姓名： | 周永权 焦李成 |
| |
| 作者单位： | 1. 西安电子科技大学智能信息处理研究所,陕西西安 710071;2. 广西民族学院计算机与信息科学学院,广西南宁 530006 |
| |
| 基金项目： | 国家自然科学基金，广西高校百名中青年学科带头人计划 |
| |
| 摘 要： | 针对高属性维稀疏数据聚类问题,定义了模糊取大逻辑神经元,给出一种新的单层离散型回归逻辑神经网络模型,由稀疏特征差异度组成的相似阵作为该网络的初始权矩阵,通过单层离散型回归逻辑神经网络学习算法,可求出相似矩阵的等价阵,根据等价阵,给定不同的阈值,可动态地、有效地实现对高属性维稀疏数据的归并,使得聚类结果更符合实际情况,聚类质量较高.相比同类聚类算法,它具有学习、修正和应变功能,适用于大规模稀疏数据库和稀疏数据仓库的聚类分析.
|
| 关 键 词： | 稀疏特征 差异度 逻辑回归神经网络 动态聚类法 学习算法 |
| 文章编号： | 0372-2112(2004)08-1342-04 |
| 收稿时间： | 2003-07-15 |
High Attribute Dimensional Sparse Clustering Recurrent Logical Neural Networks Model and Learning Algorithm |
| |
| ZHOU Yong-quan ,JIAO Li-cheng.High Attribute Dimensional Sparse Clustering Recurrent Logical Neural Networks Model and Learning Algorithm[J].Acta Electronica Sinica,2004,32(8):1342-1345. |
| |
| Authors: | ZHOU Yong-quan JIAO Li-cheng |
| |
| Affiliation: | 1. Institute of Intelligence Information Processing,Xidian University,Xi'an,Shaanxi 710071,China;2. College of Computer and Information Science,Guangxi University for Nationalities,Nanning,Guangxi 530006,China |
| |
| Abstract: | The concepts of high attribute dimensional information system and fuzzy logical neuron are firstly proposed,and a kind of new single layer discrete recurrent neural network model with n input and n output is designed.The learning of the weight matrix be done by means of finding weight matrix of equal matrix,and elements of initials weight matrix consist of sparse feature difference degree,so a sparse feature difference degree is given.It can realize dynamic state and effective classification,and the algorithm has important use in high attribute dimensional data mining and clustering analysis. |
| |
| Keywords: | sparse feature difference degree logical recurrent neural network dynamic state clustering learning algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《电子学报》浏览原始摘要信息 |
|
点击此处可从《电子学报》下载全文 |
