DeFed-GMM-DaDiL enables stable decentralized domain adaptation by approximating client GMMs through shared learnable atoms and labeled Wasserstein barycenters, reconstructing missing classes competitively.
Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS) , pages=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DeFed-GMM-DaDiL: A Decentralized Federated Framework for Domain Adaptation
DeFed-GMM-DaDiL enables stable decentralized domain adaptation by approximating client GMMs through shared learnable atoms and labeled Wasserstein barycenters, reconstructing missing classes competitively.