基于全局融合的多核概念分解算法

doi:10.11772/j.issn.1001-9081.2018081817

计算机应用 ›› 2019, Vol. 39 ›› Issue (4): 1021-1026.DOI: 10.11772/j.issn.1001-9081.2018081817

基于全局融合的多核概念分解算法

李飞^1,2, 杜亮^1,2,3, 任超宏^1,2

1. 山西大学计算机与信息技术学院, 太原 030006;
2. 山西大学大数据科学与产业研究院, 太原 030006;
3. 计算智能与中文信息处理教育部重点实验室(山西大学), 太原 030006

收稿日期:2018-09-13 修回日期:2018-11-21 发布日期:2019-04-10 出版日期:2019-04-10
通讯作者: 杜亮
作者简介:李飞(1993-),男,湖北黄冈人,硕士研究生,CCF会员,主要研究方向:数据挖掘、机器学习;杜亮(1985-),男,山西晋中人,讲师,博士,CCF会员,主要研究方向:数据挖掘、机器学习;任超宏(1994-),男,山西朔州人,硕士研究生,主要研究方向:数据挖掘、机器学习。
基金资助:
国家自然科学基金资助项目（61502289）。

Multiple kernel concept factorization algorithm based on global fusion

LI Fei^1,2, DU Liang^1,2,3, REN Chaohong^1,2

1. School of Computer and Information Technology, Shanxi University, Taiyuan Shanxi 030006, China;
2. Institute of Big Data Science and Industry, Shanxi University, Taiyuan Shanxi 030006, China;
3. Key Laboratory of Computational Intelligence and Chinese Information Processing, Ministry of Education(Shanxi University), Taiyuan Shanxi 030006, China

Received:2018-09-13 Revised:2018-11-21 Online:2019-04-10 Published:2019-04-10
Supported by:
This work is partially supported by the National Natural Science Foundation of China (61502289).

摘要/Abstract

摘要： 非负矩阵分解（NMF）算法仅能用于对原始非负数据寻找低秩近似，而概念分解（CF）算法将矩阵分解模型扩展到单个非线性核空间，提升了矩阵分解算法的学习能力和普适性。针对无监督环境下概念分解面临的如何设计或选择合适核函数这一问题，提出基于全局融合的多核概念分解（GMKCF）算法。同时输入多种候选核函数，在概念分解框架下基于全局线性权重融合对它们进行学习，以得出质量高稳定性好的聚类结果，并解决概念分解模型面临核函数选择的问题。采用交替迭代的方法对新模型进行求解，证明了算法的收敛性。将该算法与基于核的K-均值（KKM）、谱聚类（SC）、KCF（Kernel Concept Factorization）、Coreg（Co-regularized multi-view spectral clustering）、RMKKM（Robust Multiple KKM）在多个真实数据库上的实验结果表明，该算法在数据聚类方面优于对比算法。

关键词: 多核学习, 概念分解, 矩阵分解, 多核聚类, 全局融合

Abstract: Non-negative Matrix Factorization (NMF) algorithm can only be used to find low rank approximation of original non-negative data while Concept Factorization (CF) algorithm extends matrix factorization to single non-linear kernel space, improving learning ability and adaptability of matrix factorization. In unsupervised environment, to design or select proper kernel function for specific dataset, a new algorithm called Globalized Multiple Kernel CF (GMKCF) was proposed. Multiple candidate kernel functions were input in the same time and learned in the CF framework based on global linear fusion, obtaining a clustering result with high quality and stability and solving the problem of kernel function selection that the CF faced. The convergence of the proposed algorithm was verified by solving the model with alternate iteration. The experimental results on several real databases show that the proposed algorithm outperforms comparison algorithms in data clustering, such as Kernel K-Means (KKM), Spectral Clustering (SC), Kernel CF (KCF), Co-regularized multi-view spectral clustering (Coreg), and Robust Multiple KKM (RMKKM).

Key words: multiple kernel learning, Concept Factorization (CF), matrix factorization, multiple kernel clustering, global fusion

中图分类号:

TP181

李飞, 杜亮, 任超宏. 基于全局融合的多核概念分解算法[J]. 计算机应用, 2019, 39(4): 1021-1026.

LI Fei, DU Liang, REN Chaohong. Multiple kernel concept factorization algorithm based on global fusion[J]. Journal of Computer Applications, 2019, 39(4): 1021-1026.

参考文献

[1] HAN J, KAMBER M, PEI J. Data Mining:Concepts and Techniques[M]. 3rd ed. San Francisco:Margan Kaufmann, 2011:525-527.
[2] LEE D D, HSEBASTIAN S S. Learning the parts of objects by non-negative matrix factorization[J]. Nature, 1999, 401:788-791.
[3] CESARIO E, MANCO G, ORTALE R. Top-down parameter-free clustering of high-dimensional categorical data[J]. IEEE Transactions on Knowledge and Data Engineering, 2007, 19(12):1607-1624.
[4] SHI J, MALIK J. Normalized cuts and image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(8):888-905.
[5] FREY B J, DUECK D. Clustering by passing messages between data points[J]. Science, 2007, 315(5814):972-976.
[6] CAI D, HE X, HAN J, et al. Graph regularized nonnegative matrix factorization for data representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(8):1548-1560.
[7] BISHOP C M. Pattern Recognition and Machine Learning[M]. 2nd ed. New York:Springer, 2010:291-292.
[8] HOYER P O. Non-negative matrix factorization with sparseness constraints[EB/OL].[2018-05-10]. https://arxiv.org/abs/cs/0408058.
[9] DU L, LI X, SHEN Y. Robust nonnegative matrix factorization via half-quadratic minimization[C]//Proceedings of the 2012 IEEE 12th International Conference on Data Mining. Piscataway, NJ:IEEE, 2012:201-210.
[10] XU W, GONG Y. Document clustering by concept factorization[C]//SIGIR 2004:Proceedings of the 27th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. New York:ACM, 2004:202-209.
[11] CAI D, HE X, HAN J. Locally consistent concept factorization for document clustering[J]. IEEE Transactions on Knowledge and Data Engineering, 2011, 23(6):902-913.
[12] PEI X, CHEN C, GONG W. Concept factorization with adaptive neighbors for document clustering[J]. IEEE Transactions on Neural Networks and Learning Systems, 2018, 29(2):343-352.
[13] LEE D D, SEUNG H S. Algorithms for non-negative matrix factorization[EB/OL].[2018-05-10]. http://papers.nips.cc/paper/1861-algorithms-for-non-negative-matrix-factorization.pdf.
[14] KUMAR A, RAI P, DAUMÉ H. Co-regularized multi-view spectral clustering[EB/OL].[2018-05-10]. http://www.cs.utah.edu/~piyush/recent/spectral-nips11.pdf.
[15] DU L, ZHOU P, SHI L, et al. Robust multiple kernel k-means clustering using L₂₁-norm[C]//Proceedings of the Twenty-Fourth International Joint Conference on Artificial Intelligence. Menlo Park, CA:AAAI Press, 2015:3476-3482.
[16] LI X, SHEEN X, SHU Z, et al. Graph regularized multilayer concept factorization for data representation[J]. Neurocomputing, 2017, 238:139-151.
[17] ZHAN K, SHI J, WANG J, et al. Adaptive structure concept factorization for multiview clustering[J]. Neural Computation, 2018, 30(2):1080-1103.
[18] SHU Z, WU X, HUANG P, et al. Multiple graph regularized concept factorization with adaptive weights[J]. IEEE Access, 2018, 6:64938-64945.
[19] MA S, ZHANG L, HU E, et al. Self-representative manifold concept factorization with adaptive neighbors for clustering[C]//IJCAI 2018:Proceedings of the 27th International Joint Conference on Artificial Intelligence. Menlo Park, CA:AAAI Press, 2018:2539-2545.
[20] KUMAR A, RAI P, DAUMÉ H, Ⅲ. Co-regularized multi-view spectral clustering[C]//NIPS 2011:Proceedings of the 24th International Conference on Neural Information Processing Systems. New York:ACM, 2011:1413-1421.
[21] YAN W, ZHANG B, MA S, et al. A novel regularized concept factorization for document clustering[J]. Knowledge-based Systems, 2017, 135:147-158.

基于全局融合的多核概念分解算法

Multiple kernel concept factorization algorithm based on global fusion

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	徐则林, 杨敏, 陈勐. 融合空间和文本信息的兴趣点类别表征模型[J]. 《计算机应用》唯一官方网站, 2023, 43(8): 2456-2461.
[2]	周寅莹, 周允升, 余敦辉, 孙军. 基于消极相似性的自适应社会化推荐[J]. 《计算机应用》唯一官方网站, 2023, 43(8): 2439-2447.
[3]	谭钰, 王小琴, 蓝如师, 刘振丙, 罗笑南. 基于判别性矩阵分解的多标签跨模态哈希检索[J]. 《计算机应用》唯一官方网站, 2023, 43(5): 1349-1354.
[4]	李振亮, 李波. 基于矩阵分解的卷积神经网络改进方法[J]. 《计算机应用》唯一官方网站, 2023, 43(3): 685-691.
[5]	任奇泽, 贾洪杰, 陈东宇. 融合局部结构学习的大规模子空间聚类算法[J]. 《计算机应用》唯一官方网站, 2023, 43(12): 3747-3754.
[6]	包永春, 张建臣, 杜守信, 张军军. 基于非负矩阵分解与稀疏表示的多标签分类算法[J]. 《计算机应用》唯一官方网站, 2022, 42(5): 1375-1382.
[7]	祁祥洲, 邢红杰. 基于中心核对齐的多核单类支持向量机[J]. 《计算机应用》唯一官方网站, 2022, 42(2): 349-356.
[8]	周寅莹, 章梦怡, 余敦辉, 朱明. 融合信任隐含相似度与评分相似度的社会化推荐[J]. 《计算机应用》唯一官方网站, 2022, 42(12): 3671-3678.
[9]	杨悦, 王士同. 基于随机特征映射的四层多核学习方法[J]. 《计算机应用》唯一官方网站, 2022, 42(1): 16-25.
[10]	王锦凯, 贾旭. 基于迁移孪生非负矩阵分解的静脉识别算法[J]. 计算机应用, 2021, 41(3): 898-903.
[11]	陈献, 胡丽莹, 林晓炜, 陈黎飞. 基于核非负矩阵分解的有向图聚类算法[J]. 《计算机应用》唯一官方网站, 2021, 41(12): 3447-3454.
[12]	杜汉, 龙显忠, 李云. 基于图学习正则判别非负矩阵分解的人脸识别[J]. 《计算机应用》唯一官方网站, 2021, 41(12): 3455-3461.
[13]	王梅, 许传海, 刘勇. 基于神经正切核的多核学习方法[J]. 《计算机应用》唯一官方网站, 2021, 41(12): 3462-3467.
[14]	李华, 卢桂馥, 余沁茹. 基于干净数据的流形正则化非负矩阵分解[J]. 《计算机应用》唯一官方网站, 2021, 41(12): 3492-3498.
[15]	骆锦潍, 刘杜钢, 潘微科, 明仲. 基于改进的倾向得分估计的无偏推荐模型[J]. 《计算机应用》唯一官方网站, 2021, 41(12): 3508-3514.