H-ensemble: An Information Theoretic Approach to Reliable Few-Shot Multi-Source-Free Transfer
DOI:
https://doi.org/10.1609/aaai.v38i14.29528Keywords:
ML: Transfer, Domain Adaptation, Multi-Task Learning, ML: Deep Learning Theory, ML: Ensemble Methods, ML: Information Theory, ML: Optimization, ML: Transparent, Interpretable, Explainable MLAbstract
Multi-source transfer learning is an effective solution to data scarcity by utilizing multiple source tasks for the learning of the target task. However, access to source data and model details is limited in the era of commercial models, giving rise to the setting of multi-source-free (MSF) transfer learning that aims to leverage source domain knowledge without such access. As a newly defined problem paradigm, MSF transfer learning remains largely underexplored and not clearly formulated. In this work, we adopt an information theoretic perspective on it and propose a framework named H-ensemble, which dynamically learns the optimal linear combination, or ensemble, of source models for the target task, using a generalization of maximal correlation regression. The ensemble weights are optimized by maximizing an information theoretic metric for transferability. Compared to previous works, H-ensemble is characterized by: 1) its adaptability to a novel and realistic MSF setting for few-shot target tasks, 2) theoretical reliability, 3) a lightweight structure easy to interpret and adapt. Our method is empirically validated by ablation studies, along with extensive comparative analysis with other task ensemble and transfer learning methods. We show that the H-ensemble can successfully learn the optimal task ensemble, as well as outperform prior arts.Downloads
Published
2024-03-24
How to Cite
Wu, Y., Wang, J., Wang, W., & Li, Y. (2024). H-ensemble: An Information Theoretic Approach to Reliable Few-Shot Multi-Source-Free Transfer. Proceedings of the AAAI Conference on Artificial Intelligence, 38(14), 15970-15978. https://doi.org/10.1609/aaai.v38i14.29528
Issue
Section
AAAI Technical Track on Machine Learning V