AAAI 2025technical0 citations
Hypernetwork Approach to Bayesian MAML (Student Abstract)
Piotr Borycki, Piotr Kubacki, Marcin Przewięźlikowski, Tomasz Kuśmierczyk, Jacek Tabor, Przemysław Spurek
Abstract
The main goal of Few-Shot learning algorithms is to enable learning from small amounts of data. One of the most popular and elegant Few-Shot learning approaches is Model-Agnostic Meta-Learning (MAML). In this paper, we propose a novel framework for Bayesian MAML called BH-MAML, which employs Hypernetworks for weight updates. It learns the universal weights point-wise, but a probabilistic structure is added when adapted for specific tasks. In such a framework, we can use simple Gaussian distributions or more complicated posteriors induced by Continuous Normalizing Flows.
BibTeX
@article{Borycki_Kubacki_Przewięźlikowski_Kuśmierczyk_Tabor_Spurek_2025, title={Hypernetwork Approach to Bayesian MAML (Student Abstract)}, volume={39}, url={https://ojs.aaai.org/index.php/AAAI/article/view/35239}, DOI={10.1609/aaai.v39i28.35239}, abstractNote={The main goal of Few-Shot learning algorithms is to enable
learning from small amounts of data. One of the most popular
and elegant Few-Shot learning approaches is Model-Agnostic
Meta-Learning (MAML). In this paper, we propose a novel
framework for Bayesian MAML called BH-MAML, which
employs Hypernetworks for weight updates. It learns the
universal weights point-wise, but a probabilistic structure is
added when adapted for specific tasks. In such a framework,
we can use simple Gaussian distributions or more complicated posteriors induced by Continuous Normalizing Flows.}, number={28}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Borycki, Piotr and Kubacki, Piotr and Przewięźlikowski, Marcin and Kuśmierczyk, Tomasz and Tabor, Jacek and Spurek, Przemysław}, year={2025}, month={Apr.}, pages={29325-29327} }