English

Task Weighting in Meta-learning with Trajectory Optimisation

Machine Learning 2023-01-05 v1 Artificial Intelligence

Abstract

Developing meta-learning algorithms that are un-biased toward a subset of training tasks often requires hand-designed criteria to weight tasks, potentially resulting in sub-optimal solutions. In this paper, we introduce a new principled and fully-automated task-weighting algorithm for meta-learning methods. By considering the weights of tasks within the same mini-batch as an action, and the meta-parameter of interest as the system state, we cast the task-weighting meta-learning problem to a trajectory optimisation and employ the iterative linear quadratic regulator to determine the optimal action or weights of tasks. We theoretically show that the proposed algorithm converges to an ϵ0\epsilon_{0}-stationary point, and empirically demonstrate that the proposed approach out-performs common hand-engineering weighting methods in two few-shot learning benchmarks.

Keywords

Cite

@article{arxiv.2301.01400,
  title  = {Task Weighting in Meta-learning with Trajectory Optimisation},
  author = {Cuong Nguyen and Thanh-Toan Do and Gustavo Carneiro},
  journal= {arXiv preprint arXiv:2301.01400},
  year   = {2023}
}

Comments

Revision after a peer review from JMLR

R2 v1 2026-06-28T08:01:51.246Z