We introduce a new and rigorously-formulated PAC-Bayes meta-learning algorithm that solves few-shot learning. Our proposed method extends the PAC-Bayes framework from a single task setting to the meta-learning multiple task setting to upper-bound the error evaluated on any, even unseen, tasks and samples. We also propose a generative-based approach to estimate the posterior of task-specific model parameters more expressively compared to the usual assumption based on a multivariate normal distribution with a diagonal covariance matrix. We show that the models trained with our proposed meta-learning algorithm are well calibrated and accurate, with state-of-the-art calibration and classification results on few-shot classification (mini-ImageNet and tiered-ImageNet) and regression (multi-modal task-distribution regression) benchmarks.
@article{arxiv.2003.02455,
title = {PAC-Bayes meta-learning with implicit task-specific posteriors},
author = {Cuong Nguyen and Thanh-Toan Do and Gustavo Carneiro},
journal= {arXiv preprint arXiv:2003.02455},
year = {2021}
}
Comments
Add background and directly specify meta-learning as a bi-level optimisation