Robust spectral embedded bilateral orthogonal concept factorization for clustering
Concept factorization (CF), unlike nonnegative matrix factorization (NMF), can handle data with negative values by approximating the original data with two low-dimensional nonnegative matrices and itself. Nevertheless, existing CF-based methods continue to suffer from the two issues specified as fol...
| Main Authors: | , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/180168 |
| _version_ | 1811681582157660160 |
|---|---|
| author | Yang, Ben Wu, Jinghan Zhou, Yu Zhang, Xuetao Lin, Zhiping Nie, Feiping Chen, Badong |
| author2 | School of Electrical and Electronic Engineering |
| author_facet | School of Electrical and Electronic Engineering Yang, Ben Wu, Jinghan Zhou, Yu Zhang, Xuetao Lin, Zhiping Nie, Feiping Chen, Badong |
| author_sort | Yang, Ben |
| collection | NTU |
| description | Concept factorization (CF), unlike nonnegative matrix factorization (NMF), can handle data with negative values by approximating the original data with two low-dimensional nonnegative matrices and itself. Nevertheless, existing CF-based methods continue to suffer from the two issues specified as follows: (1) Their effectiveness is reduced by the high degree of factorization freedom and the two-stage mismatch between factorization and category acquisition, and (2) their robustness drops significantly when dealing with complex noise. In response to the aforementioned issues, we propose a robust spectral-embedded bilateral orthogonal concept factorization (RSOCF) model for clustering. It constrains the factor matrices as orthogonal matrices to decrease the freedom and obtain samples’ categories directly after factorization, which can significantly improve clustering effectiveness. Moreover, correntropy is introduced into RSOCF to improve its robustness to complex noise. To optimize the non-convex RSOCF model, a half-quadratic-based algorithm is devised. Numerous experiments demonstrate that RSOCF surpasses other state-of-the-art methods in terms of clustering effectiveness and robustness. |
| first_indexed | 2024-10-01T03:43:14Z |
| format | Journal Article |
| id | ntu-10356/180168 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T03:43:14Z |
| publishDate | 2024 |
| record_format | dspace |
| spelling | ntu-10356/1801682024-09-23T01:24:40Z Robust spectral embedded bilateral orthogonal concept factorization for clustering Yang, Ben Wu, Jinghan Zhou, Yu Zhang, Xuetao Lin, Zhiping Nie, Feiping Chen, Badong School of Electrical and Electronic Engineering Engineering Concept factorization Spectral embedding Concept factorization (CF), unlike nonnegative matrix factorization (NMF), can handle data with negative values by approximating the original data with two low-dimensional nonnegative matrices and itself. Nevertheless, existing CF-based methods continue to suffer from the two issues specified as follows: (1) Their effectiveness is reduced by the high degree of factorization freedom and the two-stage mismatch between factorization and category acquisition, and (2) their robustness drops significantly when dealing with complex noise. In response to the aforementioned issues, we propose a robust spectral-embedded bilateral orthogonal concept factorization (RSOCF) model for clustering. It constrains the factor matrices as orthogonal matrices to decrease the freedom and obtain samples’ categories directly after factorization, which can significantly improve clustering effectiveness. Moreover, correntropy is introduced into RSOCF to improve its robustness to complex noise. To optimize the non-convex RSOCF model, a half-quadratic-based algorithm is devised. Numerous experiments demonstrate that RSOCF surpasses other state-of-the-art methods in terms of clustering effectiveness and robustness. This work was supported in part by the National Natural Science Foundation of China (No. 62088102), the National Postdoctoral Program for Innovative Talents (No. BX20230283), the Project funded by China Postdoctoral Science Foundation (No. 2023M742791), the Natural Science Basic Research Program of Shaanxi Province (No. 2023- JC-YB-486, No. 2024JC-YBQN-0658), and the Fundamental Research Funds for the Central Universities (No. xzy012023137). 2024-09-23T01:24:40Z 2024-09-23T01:24:40Z 2024 Journal Article Yang, B., Wu, J., Zhou, Y., Zhang, X., Lin, Z., Nie, F. & Chen, B. (2024). Robust spectral embedded bilateral orthogonal concept factorization for clustering. Pattern Recognition, 150, 110308-. https://dx.doi.org/10.1016/j.patcog.2024.110308 0031-3203 https://hdl.handle.net/10356/180168 10.1016/j.patcog.2024.110308 2-s2.0-85184771358 150 110308 en Pattern Recognition © 2024 Elsevier Ltd. All rights reserved. |
| spellingShingle | Engineering Concept factorization Spectral embedding Yang, Ben Wu, Jinghan Zhou, Yu Zhang, Xuetao Lin, Zhiping Nie, Feiping Chen, Badong Robust spectral embedded bilateral orthogonal concept factorization for clustering |
| title | Robust spectral embedded bilateral orthogonal concept factorization for clustering |
| title_full | Robust spectral embedded bilateral orthogonal concept factorization for clustering |
| title_fullStr | Robust spectral embedded bilateral orthogonal concept factorization for clustering |
| title_full_unstemmed | Robust spectral embedded bilateral orthogonal concept factorization for clustering |
| title_short | Robust spectral embedded bilateral orthogonal concept factorization for clustering |
| title_sort | robust spectral embedded bilateral orthogonal concept factorization for clustering |
| topic | Engineering Concept factorization Spectral embedding |
| url | https://hdl.handle.net/10356/180168 |
| work_keys_str_mv | AT yangben robustspectralembeddedbilateralorthogonalconceptfactorizationforclustering AT wujinghan robustspectralembeddedbilateralorthogonalconceptfactorizationforclustering AT zhouyu robustspectralembeddedbilateralorthogonalconceptfactorizationforclustering AT zhangxuetao robustspectralembeddedbilateralorthogonalconceptfactorizationforclustering AT linzhiping robustspectralembeddedbilateralorthogonalconceptfactorizationforclustering AT niefeiping robustspectralembeddedbilateralorthogonalconceptfactorizationforclustering AT chenbadong robustspectralembeddedbilateralorthogonalconceptfactorizationforclustering |