MeTa Learning-Based Optimization of Unsupervised Domain Adaptation Deep Networks

Lin, Hsiau-Wen; Ho, Trang-Thi; Tu, Ching-Ting; Lin, Hwei-Jen; Yu, Chen-Hsiang

doi:10.3390/math13020226

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

MeTa Learning-Based Optimization of Unsupervised Domain Adaptation Deep Networks

by

Hsiau-Wen Lin

^1,*,

Trang-Thi Ho

²,

Ching-Ting Tu

^3,*,

Hwei-Jen Lin

^2,* and

Chen-Hsiang Yu

⁴

¹

Department of Information Management, Chihlee University of Technology, Taipei 220305, Taiwan

²

Department of Computer Science and Information Engineering, Tamkang University, Taipei 251301, Taiwan

³

Department of Applied Mathematics, National Chung Hsing University, Taichung 402202, Taiwan

⁴

Multidisciplinary Graduate Engineering, College of Engineering, Northeastern University, Boston, MA 02115, USA

^*

Authors to whom correspondence should be addressed.

Mathematics 2025, 13(2), 226; https://doi.org/10.3390/math13020226

Submission received: 12 December 2024 / Revised: 5 January 2025 / Accepted: 6 January 2025 / Published: 10 January 2025

(This article belongs to the Special Issue Advances in Intelligent Computing, Machine Learning and Pattern Recognition)

Download Versions Notes

Abstract

This paper introduces a novel unsupervised domain adaptation (UDA) method, MeTa Discriminative Class-Wise MMD (MCWMMD), which combines meta-learning with a Class-Wise Maximum Mean Discrepancy (MMD) approach to enhance domain adaptation. Traditional MMD methods align overall distributions but struggle with class-wise alignment, reducing feature distinguishability. MCWMMD incorporates a meta-module to dynamically learn a deep kernel for MMD, improving alignment accuracy and model adaptability. This meta-learning technique enhances the model’s ability to generalize across tasks by ensuring domain-invariant and class-discriminative feature representations. Despite the complexity of the method, including the need for meta-module training, it presents a significant advancement in UDA. Future work will explore scalability in diverse real-world scenarios and further optimize the meta-learning framework. MCWMMD offers a promising solution to the persistent challenge of domain adaptation, paving the way for more adaptable and generalizable deep learning models.

Keywords: unsupervised domain adaptation; maximum mean discrepancy (MMD); discriminative class-wise MMD (DCWMMD); meta-learning; deep kernel; feature distributions; domain shift; transfer learning

Share and Cite

MDPI and ACS Style

Lin, H.-W.; Ho, T.-T.; Tu, C.-T.; Lin, H.-J.; Yu, C.-H. MeTa Learning-Based Optimization of Unsupervised Domain Adaptation Deep Networks. Mathematics 2025, 13, 226. https://doi.org/10.3390/math13020226

AMA Style

Lin H-W, Ho T-T, Tu C-T, Lin H-J, Yu C-H. MeTa Learning-Based Optimization of Unsupervised Domain Adaptation Deep Networks. Mathematics. 2025; 13(2):226. https://doi.org/10.3390/math13020226

Chicago/Turabian Style

Lin, Hsiau-Wen, Trang-Thi Ho, Ching-Ting Tu, Hwei-Jen Lin, and Chen-Hsiang Yu. 2025. "MeTa Learning-Based Optimization of Unsupervised Domain Adaptation Deep Networks" Mathematics 13, no. 2: 226. https://doi.org/10.3390/math13020226

APA Style

Lin, H.-W., Ho, T.-T., Tu, C.-T., Lin, H.-J., & Yu, C.-H. (2025). MeTa Learning-Based Optimization of Unsupervised Domain Adaptation Deep Networks. Mathematics, 13(2), 226. https://doi.org/10.3390/math13020226

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

MeTa Learning-Based Optimization of Unsupervised Domain Adaptation Deep Networks

Abstract

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI