Cross-Domain Relation Adaptation

Ido Kessler, Omri Lifshitz, Sagie Benaim, Lior Wolf
Proceedings of the 15th Asian Conference on Machine Learning, PMLR 222:630-645, 2024.

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.

Cite this Paper

BibTeX
@InProceedings{pmlr-v222-kessler24a, title = {Cross-Domain Relation Adaptation}, author = {Kessler, Ido and Lifshitz, Omri and Benaim, Sagie and Wolf, Lior}, booktitle = {Proceedings of the 15th Asian Conference on Machine Learning}, pages = {630--645}, year = {2024}, editor = {Yanıkoğlu, Berrin and Buntine, Wray}, volume = {222}, series = {Proceedings of Machine Learning Research}, month = {11--14 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v222/kessler24a/kessler24a.pdf}, url = {https://proceedings.mlr.press/v222/kessler24a.html}, 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.} }
Endnote
%0 Conference Paper %T Cross-Domain Relation Adaptation %A Ido Kessler %A Omri Lifshitz %A Sagie Benaim %A Lior Wolf %B Proceedings of the 15th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Berrin Yanıkoğlu %E Wray Buntine %F pmlr-v222-kessler24a %I PMLR %P 630--645 %U https://proceedings.mlr.press/v222/kessler24a.html %V 222 %X 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.
APA
Kessler, I., Lifshitz, O., Benaim, S. & Wolf, L.. (2024). Cross-Domain Relation Adaptation. Proceedings of the 15th Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 222:630-645 Available from https://proceedings.mlr.press/v222/kessler24a.html.

Related Material