TechRxiv
1/1
2 files

Continual Learning with Quasi-Newton Methods

preprint
posted on 25.09.2021, 02:49 by Steven Vander EecktSteven Vander Eeckt, Hugo Van hamme
In this paper, we propose CSQN, a new Continual Learning (CL) method which considers Quasi-Newton methods, more specifically, Sampled Quasi-Newton methods, to extend EWC.
EWC uses a Bayesian framework to estimate which parameters are important to previous tasks, and it punishes changes made to these parameters. However, it assumes that parameters are independent, as it does not consider interactions between parameters. With CSQN, we aim to overcome this.

History

Email Address of Submitting Author

steven.vandereeckt@esat.kuleuven.be

Submitting Author's Institution

KU Leuven

Submitting Author's Country

Belgium