Meta-strategy for Learning Tuning Parameters with Guarantees

TL;DR

This paper introduces a meta-learning strategy that adaptively tunes parameters for online learning algorithms like OGA and EWA, providing guarantees and improving performance across tasks.

Contribution

It proposes a regret-based meta-strategy to learn initialization, step size, and prior in online algorithms with theoretical guarantees.

Findings

The strategy effectively learns parameters from past tasks.

It provides regret bounds for the meta-learning approach.

Meta-learning can outperform isolated task learning.

Abstract

Online learning methods, like the online gradient algorithm (OGA) and exponentially weighted aggregation (EWA), often depend on tuning parameters that are difficult to set in practice. We consider an online meta-learning scenario, and we propose a meta-strategy to learn these parameters from past tasks. Our strategy is based on the minimization of a regret bound. It allows to learn the initialization and the step size in OGA with guarantees. It also allows to learn the prior or the learning rate in EWA. We provide a regret analysis of the strategy. It allows to identify settings where meta-learning indeed improves on learning each task in isolation.