.png){kind=icon}
返回
FGCH: a fast and grid based clustering algorithm for hybrid data stream
.png){kind=icon}
PRE
AI
DOI：10.1007/s10489-018-1324-x.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Streaming large volumes of data has a wide range of real-world applications, e.g., video flows, internet calls, and online games etc. Thus, fast and real-time data stream processing is important. Traditionally, data clustering algorithms are efficient and effective to mine information from large data. However, they are mostly not suitable for online data stream clustering. Therefore, in this work, we propose a novel fast and grid based clustering algorithm for hybrid data stream (FGCH). Specifically, we have made the following main contributions: 1), we develop a non-uniform attenuation model to enhance the resistance to noise; 2), we propose a similarity calculation method for hybrid data, which can calculate the similarity more efficiently and accurately; and 3), we present a novel clustering center fast determination algorithm (CCFD), which can automatically determine the number, center, and radius of clusters. Our technique is compared with several state-of-art clustering algorithms. The experimental results show that our technique can achieve more than better clustering accuracy on average. Meanwhile, the running time is shorter compared with the closest algorithm.
Keyword:
Data stream
Clustering analysis
Non-uniform attenuation
Grid clustering
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
3.5
论文数:
7.6K
被引数:
1.7W
机构
引用论文
An improved k-prototypes clustering algorithm for mixed numeric and categorical data.png)
NEUROCOMPUTING
NEUROCOMPUTINGIF6.5
PREAI
分享收藏
Automatic Pearl Classification Machine Based on a Multistream Convolutional Neural Network基于多流卷积神经网络的珍珠自动分类机
PREAI
分享收藏
查看更多内容