Achieving Forgetting Prevention and Knowledge Transfer in Continual Learning

TL;DR

This paper introduces CTR, a novel continual learning model that effectively balances catastrophic forgetting prevention and knowledge transfer, especially leveraging pre-trained models for improved performance.

Contribution

The paper proposes CTR, a new model that enhances continual learning by addressing both forgetting and transfer, and effectively utilizing pre-trained models.

Findings

CTR outperforms existing methods in balancing CF and KT

Pre-trained models significantly improve CL performance with CTR

CTR demonstrates robustness across different tasks and datasets

Abstract

Continual learning (CL) learns a sequence of tasks incrementally with the goal of achieving two main objectives: overcoming catastrophic forgetting (CF) and encouraging knowledge transfer (KT) across tasks. However, most existing techniques focus only on overcoming CF and have no mechanism to encourage KT, and thus do not do well in KT. Although several papers have tried to deal with both CF and KT, our experiments show that they suffer from serious CF when the tasks do not have much shared knowledge. Another observation is that most current CL methods do not use pre-trained models, but it has been shown that such models can significantly improve the end task performance. For example, in natural language processing, fine-tuning a BERT-like pre-trained language model is one of the most effective approaches. However, for CL, this approach suffers from serious CF. An interesting question is how to make the best use of pre-trained models for CL. This paper proposes a novel model called CTR to solve these problems. Our experimental results demonstrate the effectiveness of CTR