Structure preserved ordinal unsupervised domain adaptation
Unsupervised domain adaptation (UDA) aims to transfer the knowledge from labeled source domain to unlabeled target domain. The main challenge of UDA stems from the domain shift between the source and target domains. Currently, in the discrete classification problems, most existing UDA methods usuall...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIMS Press
2024-11-01
|
| Series: | Electronic Research Archive |
| Subjects: | |
| Online Access: | https://www.aimspress.com/article/doi/10.3934/era.2024295 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832590780983672832 |
|---|---|
| author | Qing Tian Canyu Sun |
| author_facet | Qing Tian Canyu Sun |
| author_sort | Qing Tian |
| collection | DOAJ |
| description | Unsupervised domain adaptation (UDA) aims to transfer the knowledge from labeled source domain to unlabeled target domain. The main challenge of UDA stems from the domain shift between the source and target domains. Currently, in the discrete classification problems, most existing UDA methods usually adopt the distribution alignment strategy while enforcing unstable instances to pass through the low-density areas. However, the scenario of ordinal regression (OR) is rarely researched in UDA, and the traditional UDA methods cannot preferably handle OR since they do not preserve the order relationships in data labels, like in human age estimation. To address this issue, we proposed a structure-oriented adaptation strategy, namely, structure preserved ordinal unsupervised domain adaptation (SPODA). More specifically, on one hand, the global structure information was modeled and embedded into an auto-encoder framework via a low-rank transferred structure matrix. On the other hand, the local structure information was preserved through a weighted pair-wise strategy in the latent space. Guided by both the local and global structure information, a well-performance latent space was generated, whose geometric structure was adopted to further obtain a more discriminant ordinal regressor. To further enhance its generalization, a counterpart of SPODA with deep architecture was developed. Finally, extensive experiments indicated that in addressing the OR problem, SPODA was more effective and advanced than existing related domain adaptation methods. |
| format | Article |
| id | doaj-art-dc73389984584688a7da2ed6fad4a8ef |
| institution | Kabale University |
| issn | 2688-1594 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | AIMS Press |
| record_format | Article |
| series | Electronic Research Archive |
| spelling | doaj-art-dc73389984584688a7da2ed6fad4a8ef2025-01-23T07:53:01ZengAIMS PressElectronic Research Archive2688-15942024-11-0132116338636310.3934/era.2024295Structure preserved ordinal unsupervised domain adaptationQing Tian0Canyu Sun1School of Software, Nanjing University of Information Science and Technology, Nanjing 210044, ChinaSchool of Software, Nanjing University of Information Science and Technology, Nanjing 210044, ChinaUnsupervised domain adaptation (UDA) aims to transfer the knowledge from labeled source domain to unlabeled target domain. The main challenge of UDA stems from the domain shift between the source and target domains. Currently, in the discrete classification problems, most existing UDA methods usually adopt the distribution alignment strategy while enforcing unstable instances to pass through the low-density areas. However, the scenario of ordinal regression (OR) is rarely researched in UDA, and the traditional UDA methods cannot preferably handle OR since they do not preserve the order relationships in data labels, like in human age estimation. To address this issue, we proposed a structure-oriented adaptation strategy, namely, structure preserved ordinal unsupervised domain adaptation (SPODA). More specifically, on one hand, the global structure information was modeled and embedded into an auto-encoder framework via a low-rank transferred structure matrix. On the other hand, the local structure information was preserved through a weighted pair-wise strategy in the latent space. Guided by both the local and global structure information, a well-performance latent space was generated, whose geometric structure was adopted to further obtain a more discriminant ordinal regressor. To further enhance its generalization, a counterpart of SPODA with deep architecture was developed. Finally, extensive experiments indicated that in addressing the OR problem, SPODA was more effective and advanced than existing related domain adaptation methods.https://www.aimspress.com/article/doi/10.3934/era.2024295unsupervised domain adaptationordinal domain adaptationstructure-oriented adaptationordinal regression |
| spellingShingle | Qing Tian Canyu Sun Structure preserved ordinal unsupervised domain adaptation Electronic Research Archive unsupervised domain adaptation ordinal domain adaptation structure-oriented adaptation ordinal regression |
| title | Structure preserved ordinal unsupervised domain adaptation |
| title_full | Structure preserved ordinal unsupervised domain adaptation |
| title_fullStr | Structure preserved ordinal unsupervised domain adaptation |
| title_full_unstemmed | Structure preserved ordinal unsupervised domain adaptation |
| title_short | Structure preserved ordinal unsupervised domain adaptation |
| title_sort | structure preserved ordinal unsupervised domain adaptation |
| topic | unsupervised domain adaptation ordinal domain adaptation structure-oriented adaptation ordinal regression |
| url | https://www.aimspress.com/article/doi/10.3934/era.2024295 |
| work_keys_str_mv | AT qingtian structurepreservedordinalunsuperviseddomainadaptation AT canyusun structurepreservedordinalunsuperviseddomainadaptation |