Timezone: »

Generalization Bounds For Meta-Learning: An Information-Theoretic Analysis
Qi CHEN · Changjian Shui · Mario Marchand

@ None

We derive a novel information-theoretic analysis of the generalization property of meta-learning algorithms. Concretely, our analysis proposes a generic understanding in both the conventional learning-to-learn framework \citep{amit2018meta} and the modern model-agnostic meta-learning (MAML) algorithms \citep{finn2017model}.Moreover, we provide a data-dependent generalization bound for the stochastic variant of MAML, which is \emph{non-vacuous} for deep few-shot learning. As compared to previous bounds that depend on the square norms of gradients, empirical validations on both simulated data and a well-known few-shot benchmark show that our bound is orders of magnitude tighter in most conditions.

Author Information

Qi CHEN (Laval University)
Changjian Shui (Université Laval)
Mario Marchand (Université Laval)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors