Meta-strategy for Learning Tuning Parameters with Guarantees
arXiv:2102.02504 · doi:10.3390/e23101257
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.
References in corpus (7)
- Theoretical Models of Learning to Learn
- Learning to learn by gradient descent by gradient descent
- A Generalization Bound for Online Variational Inference
- Temporal Variability in Implicit Online Learning
- Provable Smoothness Guarantees for Black-Box Variational Inference
- The Advantage of Conditional Meta-Learning for Biased Regularization and Fine-Tuning
- Non-exponentially weighted aggregation: regret bounds for unbounded loss functions