Title:
MULTI-VIEW CLUSTERING METHOD BASED ON CONSISTENT GRAPH LEARNING
Document Type and Number:
WIPO Patent Application WO/2022/166366
Kind Code:
A1
Abstract:
Disclosed in the present application is a multi-view clustering method based on consistent graph learning, comprising: S11. inputting an original data matrix to obtain a spectral embedding matrix; S12. calculating a similarity graph matrix and a Laplacian matrix according to the spectral embedding matrix; S13. performing spectral clustering on the calculated similarity graph matrix to obtain a spectral embedding characterization; S14. stacking an inner product of the standardized spectral embedding characterization into a third-order tensor, and using low-rank tensor characterization learning to obtain a consistent distance matrix; S15. integrating spectral embedding characterization learning and low-rank tensor characterization learning into a uniform learning framework to obtain a target function; S16. solving the obtained target function by means of an alternate iterative optimization strategy; S17. constructing a consistent similarity graph according to the solve result; and S18. performing spectral clustering on the consistent similarity graph to obtain the clustering result. The present application constructs a consistent similarity graph from a spectral embedding feature for clustering. In this low-dimensional space, noise and redundant information are effectively filtered, and therefore, the obtained similarity graph can well describe a cluster structure of the data.
Inventors:
ZHU XINZHONG (CN)
XU HUIYING (CN)
LI ZHENGLAI (CN)
TANG CHANG (CN)
ZHAO JIANMIN (CN)
XU HUIYING (CN)
LI ZHENGLAI (CN)
TANG CHANG (CN)
ZHAO JIANMIN (CN)
Application Number:
PCT/CN2021/135989
Publication Date:
August 11, 2022
Filing Date:
December 07, 2021
Export Citation:
Assignee:
UNIV ZHEJIANG NORMAL (CN)
International Classes:
G06K9/62
Foreign References:
| CN112990264A | 2021-06-18 | |||
| CN110598740A | 2019-12-20 | |||
| CN107885787A | 2018-04-06 | |||
| CN108776812A | 2018-11-09 | |||
| US20210019325A1 | 2021-01-21 |
Other References:
XIA JING, DING SHIFEI: "Self-weighted Multi-view Subspace Clustering with Low-rank Sparse Constrain", NANJING DAXUE XUEBAO. ZIRAN KEXUE - JOURNAL OF NANJING UNIVERSITY. NATURAL SCIENCE EDITION, NANJING DAXUEXUEBAO BIANWEIHUI, NANJING, CN, vol. 56, no. 6, 30 November 2020 (2020-11-30), CN , pages 862 - 869, XP055957286, ISSN: 0469-5097
Attorney, Agent or Firm:
BEIJING TIANDA INTELLECTUAL PROPERTY OFFICE (CN)
Download PDF: