Bayesian Model-Agnostic Meta-Learning
Abstract
Learning to infer Bayesian posterior from a few-shot dataset is an important step towards robust meta-learning due to the model uncertainty inherent in the problem. In this paper, we propose a novel Bayesian model-agnostic meta-learning method. The proposed method combines scalable gradient-based meta-learning with nonparametric variational inference in a principled probabilistic framework. During fast adaptation, the method is capable of learning complex uncertainty structure beyond a point estimate or a simple Gaussian approximation. In addition, a robust Bayesian meta-update mechanism with a new meta-loss prevents overfitting during meta-update. Remaining an efficient gradient-based meta-learner, the method is also model-agnostic and simple to implement. Experiment results show the accuracy and robustness of the proposed method in various tasks: sinusoidal regression, image classification, active learning, and reinforcement learning.
Cite
@article{arxiv.1806.03836,
title = {Bayesian Model-Agnostic Meta-Learning},
author = {Taesup Kim and Jaesik Yoon and Ousmane Dia and Sungwoong Kim and Yoshua Bengio and Sungjin Ahn},
journal= {arXiv preprint arXiv:1806.03836},
year = {2018}
}
Comments
First two authors contributed equally. 15 pages with appendix including experimental details. Accepted in NIPS 2018