SIA OpenIR  > 数字工厂研究室
基于谱分析的密度峰值快速聚类方法算法
Alternative TitleClustering by fast search and find of density peaks based on spectrum analyze
韩忠华1,2; 毕开元1; 司雯1; 吕哲1
Department数字工厂研究室
Source Publication计算机应用
ISSN1001-9081
2019
Volume39Issue:2Pages:409-413
Contribution Rank1
Funding Organization国家自然科学基金(61503259);辽宁省科技厅面上项目(201602608);辽宁省高等学校基本科研项目(LJZ2017015)
Keyword数据聚类 适应性 降维 密度峰值聚类 谱分析
Abstract针对密度峰值快速聚类算法对不同数据集聚类效果的差异,利用谱聚类对密度峰值快速聚类算法加以改进,提出了一种基于谱分析的密度峰值快速聚类方法(Clustering by Fast Search and Find of Density Peaks based on Spectrum Analyze,CFSFDP-SA)。首先,将高维非线性的数据集映射到低维子空间上实现降维处理,将聚类问题转化为图的最优划分问题以增强算法对数据全局结构的适应性;然后,利用CFSFDP算法对处理后的数据集进行聚类。结合此两种聚类算法各自的优势,进一步提升聚类算法的性能。经数据集检测,改进后的CFSFDP算法在不同数据集上的聚类精度均得到了提升。通过人造非线性数据集和机器学习库数据集的验证结果表明该改进算法在聚类类数和聚类精度上得到了进一步优化。实验结果表明,该方法提高了CFSFDP对原始数据集的适应性,在高维数据集的聚类精度上得到了最多14%左右的提升。
Other AbstractSince the difference of clustering effect of Clustering by Fast Search and Find of Density Peaks (CFSFDP) on datasets, using spectral clustering to improve the Clustering by Fast Search and Find of Density Peaks, a Clustering by Fast Search and Find of Density Peaks Based on Spectrum Analysis (CFSFDP-SA) was proposed. Firstly, the high-dimensional non-linear dataset was mapped onto the low-dimensional subspace to realize dimension reduction. The clustering problem was transformed into the optimal partitioning problem of the graph to enhance the adaptability of the algorithm to the global structure of the data. Then the CFSFDP algorithm was used to cluster the processed dataset. Combining the advantages of these two clustering algorithms, the performance of the clustering algorithm was further improved. Through the verification of artificial nonlinear dataset and machine learning library dataset, it is found that the improved algorithm is further optimized in clustering class and clustering precision. The experimental results show that this method improves the adaptability of CFSFDP to the original data set, has improved the clustering accuracy up to 14 % on high dimensional data sets.
Language中文
Document Type期刊论文
Identifierhttp://ir.sia.cn/handle/173321/22760
Collection数字工厂研究室
Corresponding Author毕开元
Affiliation1.沈阳建筑大学信息与控制工程学院
2.中国科学院沈阳自动化研究所数字工厂研究室
Recommended Citation
GB/T 7714
韩忠华,毕开元,司雯,等. 基于谱分析的密度峰值快速聚类方法算法[J]. 计算机应用,2019,39(2):409-413.
APA 韩忠华,毕开元,司雯,&吕哲.(2019).基于谱分析的密度峰值快速聚类方法算法.计算机应用,39(2),409-413.
MLA 韩忠华,et al."基于谱分析的密度峰值快速聚类方法算法".计算机应用 39.2(2019):409-413.
Files in This Item: Download All
File Name/Size DocType Version Access License
基于谱分析的密度峰值快速聚类算法.pdf(2686KB)期刊论文作者接受稿开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[韩忠华]'s Articles
[毕开元]'s Articles
[司雯]'s Articles
Baidu academic
Similar articles in Baidu academic
[韩忠华]'s Articles
[毕开元]'s Articles
[司雯]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[韩忠华]'s Articles
[毕开元]'s Articles
[司雯]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 基于谱分析的密度峰值快速聚类算法.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.