RTRA: Rapid Training of Regularization-based Approaches in Continual Learning

TL;DR

The paper introduces RTRA, a novel modification to EWC using Natural Gradient, which enhances training efficiency in continual learning without compromising test accuracy.

Contribution

RTRA is a new approach that improves regularization-based continual learning methods by integrating Natural Gradient optimization, outperforming existing methods like EWC.

Findings

RTRA outperforms EWC on the iFood251 dataset.

RTRA maintains test accuracy while improving training efficiency.

The approach effectively mitigates catastrophic forgetting.

Abstract

Catastrophic forgetting(CF) is a significant challenge in continual learning (CL). In regularization-based approaches to mitigate CF, modifications to important training parameters are penalized in subsequent tasks using an appropriate loss function. We propose the RTRA, a modification to the widely used Elastic Weight Consolidation (EWC) regularization scheme, using the Natural Gradient for loss function optimization. Our approach improves the training of regularization-based methods without sacrificing test-data performance. We compare the proposed RTRA approach against EWC using the iFood251 dataset. We show that RTRA has a clear edge over the state-of-the-art approaches.