Target Layer Regularization for Continual Learning Using Cramer-Wold Generator

Marcin Mazur, Łukasz Pustelnik, Szymon Knop, Patryk Pagacz, Przemysław Spurek

Keywords: Continual learning, Cramer-Wold distance, regularization

Abstract: We propose an effective regularization strategy (CW-TaLaR) for solving continual learning problems. It uses a penalizing term expressed by the Cramer-Wold distance between two probability distributions defined on a target layer of an underlying neural network that is shared by all tasks, and the simple architecture of the Cramer-Wold generator for modeling output data representation. Our strategy preserves target layer distribution while learning a new task but does not require remembering previous tasks’ datasets. We perform experiments involving several common supervised frameworks, which prove the competitiveness of the CWTaLaR method in comparison to a few existing state-of-the-art continual learning models.

One-sentence Summary: We introduce a novel CL strategy (CW-TaLaR), which is based on the Cramer-Wold distance, and compare it with EWC, SI and MAS.