Arbitrary Order Metalearning With Simple Populationbased Evolution
ASecurity
**arXiv ID:** 2303.09478 **Authors:** Chris Lu, Sebastian Towers, Jakob Foerster **Published:** 2023-03-16T16:55:26Z **Abstract:** Meta-learning, the notion of learning to learn, enables learning systems to quickly and flexibly solve new tasks. This usually involves defining a set of outer-loop meta-parameters that are then used to update a set of inner-loop parameters. Most meta-learning approaches use complicated and computationally expensive bi-level optimisation schemes to update these me...
Installs into .claude/skills of the current project.
Sponsors
SSkills DirectorySkills Directory
Know which skills are safe — weekly.
SSkills DirectorySkills Directory
Is that skill safe to install?
Are you the author of Arbitrary Order Metalearning With Simple Populationbased Evolution?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/hiyenwong-arbitrary-order-metalearning-with-simple-populatio)
# Arbitrary Order Meta-Learning with Simple Population-Based Evolution
**arXiv ID:** 2303.09478
**Authors:** Chris Lu, Sebastian Towers, Jakob Foerster
**Published:** 2023-03-16T16:55:26Z
**Abstract:**
Meta-learning, the notion of learning to learn, enables learning systems to quickly and flexibly solve new tasks. This usually involves defining a set of outer-loop meta-parameters that are then used to update a set of inner-loop parameters. Most meta-learning approaches use complicated and computationally expensive bi-level optimisation schemes to update these meta-parameters. Ideally, systems should perform multiple orders of meta-learning, i.e. to learn to learn to learn and so on, to accelerate their own learning. Unfortunately, standard meta-learning techniques are often inappropriate for these higher-order meta-parameters because the meta-optimisation procedure becomes too complicated or unstable. Inspired by the higher-order meta-learning we observe in real-world evolution, we show that using simple population-based evolution implicitly optimises for arbitrarily-high order meta-parameters. First, we theoretically prove and empirically show that population-based evolution implicitly optimises meta-parameters of arbitrarily-high order in a simple setting. We then introduce a minimal self-referential parameterisation, which in principle enables arbitrary-order meta-learning. Finally, we show that higher-order meta-learning improves performance on time series forecasting tasks.
## Skill Description
This skill is generated from the arXiv paper: Arbitrary Order Meta-Learning with Simple Population-Based Evolution (2303.09478).
## How to Use
[To be filled in by the user or by future automation]
## References
- [arXiv:2303.09478](http://arxiv.org/abs/2303.09478v1)