PAC-Bayes meta-learning with implicit task-specific posteriors
This addresses the problem of few-shot learning for AI systems needing adaptation to new tasks with limited data, though it appears incremental as an extension of existing PAC-Bayes methods.
The paper tackles few-shot learning by introducing a PAC-Bayes meta-learning algorithm that provides error bounds for unseen tasks and samples, achieving state-of-the-art calibration and classification results on benchmarks like mini-ImageNet and tiered-ImageNet.
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.