On Enforcing Better Conditioned Meta-Learning for Rapid Few-Shot Adaptation. (arXiv:2206.07260v1 [cs.LG])

Web: http://arxiv.org/abs/2206.07260

Inspired by the concept of preconditioning, we propose a novel method to
increase adaptation speed for gradient-based meta-learning methods without
incurring extra parameters. We demonstrate that recasting the optimization
problem to a non-linear least-squares formulation provides a principled way to
actively enforce a $\textit{well-conditioned}$ parameter space for
meta-learning models based on the concepts of the condition number and local
curvature. Our comprehensive evaluations show that the proposed method
significantly outperforms its unconstrained counterpart especially during
initial adaptation steps, while achieving …

arxiv learning lg meta meta-learning on