Abstract
We consider the challenge of establishing relationships between samples in distinct domains, A and B, using supervised data that captures the intrinsic relationships within each domain. In other words, we present a semi-supervised setting in which there are no labeled mixed-domain pairs of samples. Our method is derived based on a generalization bound and incorporates supervised terms for each domain, a domain confusion term on the learned features, and a consistency term for domain-specific relationships when considering mixed-domain sample pairs. Our findings showcase the efficacy of our approach in two disparate domains: (i) Predicting protein-protein interactions between viruses and hosts by modeling genetic sequences. (ii) Forecasting link connections within citation graphs using graph neural networks.
Original language | English |
---|---|
Pages (from-to) | 630-645 |
Number of pages | 16 |
Journal | Proceedings of Machine Learning Research |
Volume | 222 |
State | Published - 2023 |
Event | 15th Asian Conference on Machine Learning, ACML 2023 - Istanbul, Turkey Duration: 11 Nov 2023 → 14 Nov 2023 |
Bibliographical note
Publisher Copyright:© 2023 I. Kessler1, O. Lifshitz1, S. Benaim2 & L. Wolf1.
Keywords
- Cross-domain learning
- Graph Neural Networks
- Protein-Protein Interactions
- Unsupervised Domain Adaptation