Interval-Valued Observations-Based Multi-Source Domain Adaptation Using Fuzzy Neural Networks
- Publisher:
- IEEE
- Publication Type:
- Conference Proceeding
- Citation:
- 2023 IEEE International Conference on Fuzzy Systems (FUZZ), 2023, 00, pp. 1-6
- Issue Date:
- 2023-11-09
Embargoed
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| Interval-Valued Observations-Based Multi-Source Domain Adaptation Using Fuzzy Neural Networks.pdf | Accepted version | 440.7 kB |
Copyright Clearance Process
- Recently Added
- In Progress
- Open Access
This item is currently unavailable due to the publisher's embargo.
Unsupervised domain adaptation aims to learn a well performed classifier in one domain target by leveraging knowledge from another domain source that has different distribution compared with the previous one Although researchers have gained significant achievements in addressing the domain adaptation problems existing works hold a common assumption that both the source and target domains only contain crisp valued observations However in many real world scenarios such crisp valued observations are not always available For example interval valued data where all of the observations features are described by intervals is a common type of data Hence in this paper we focus on a more realistic problem called interval valued observations based multi source domain adaptation IMSDA where we aim to train a classifier with high classi fication accuracy on an unlabeled target domain with interval valued observations by utilizing previously acquired knowledge in multiple labeled source domains with crisp valued observations We develop a practical model called fuzzy multi adversarial training neural networks FUMAT Net which integrates fuzzy neural networks and adversarial training to solve the proposed problem Further a new fuzzy relation which can measure the correlation between the multiple source domains and the target domain is proposed to improve the performance of our model The experiment results on both synthetic and real world datasets verify the efficacy of FUMAT Net and the rationality of the proposed fuzzy relation
Please use this identifier to cite or link to this item: