Local vs Global continual learning


Loading...

Date

2025

Publication Type

Conference Paper

ETH Bibliography

yes

Citations

Altmetric

Data

Abstract

Continual learning is the problem of integrating new information in a model while retaining the knowledge acquired in the past. Despite the tangible improvements achieved in recent years, the problem of continual learning is still an open one. A better understanding of the mechanisms be hind the successes and failures of existing continual learning algorithms can unlock the development of new successful strategies. In this work, we view continual learning from the perspective of the multi-task loss approximation, and we compare two alternative strategies, namely local and global approximations. We classify existing continual learning algorithms based on the approximation used, and we assess the practical effects of this distinction in common continual learning settings. Additionally, we study optimal continual learning objectives in the case of local polynomial approx imations and we provide examples of existing algorithms implementing the optimal objectiv.

Publication status

published

Book title

Proceedings of the 3rd Conference on Lifelong Learning Agents

Volume

274

Pages / Article No.

121 - 143

Publisher

PMLR

Event

3rd Conference on Lifelong Learning Agents (CoLLAs 2024)

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Organisational unit

09462 - Hofmann, Thomas / Hofmann, Thomas check_circle
02219 - ETH AI Center / ETH AI Center

Notes

Funding

Related publications and datasets