Clustering-based Domain-Incremental Learning

TL;DR

This paper introduces a clustering-based approach for domain-incremental learning that effectively mitigates catastrophic forgetting without requiring task change information, enhancing existing methods like A-GEM and OGD.

Contribution

It proposes a novel online clustering method that enables task-agnostic continual learning, addressing a key challenge in domain-incremental settings where task boundaries are unknown.

Findings

Effective counteraction of catastrophic forgetting in domain-incremental learning

Task-agnostic versions of A-GEM and OGD outperform existing methods

Demonstrated superior performance on real datasets

Abstract

We consider the problem of learning multiple tasks in a continual learning setting in which data from different tasks is presented to the learner in a streaming fashion. A key challenge in this setting is the so-called "catastrophic forgetting problem", in which the performance of the learner in an "old task" decreases when subsequently trained on a "new task". Existing continual learning methods, such as Averaged Gradient Episodic Memory (A-GEM) and Orthogonal Gradient Descent (OGD), address catastrophic forgetting by minimizing the loss for the current task without increasing the loss for previous tasks. However, these methods assume the learner knows when the task changes, which is unrealistic in practice. In this paper, we alleviate the need to provide the algorithm with information about task changes by using an online clustering-based approach on a dynamically updated finite pool of samples or gradients. We thereby successfully counteract catastrophic forgetting in one of the hardest settings, namely: domain-incremental learning, a setting for which the problem was previously unsolved. We showcase the benefits of our approach by applying these ideas to projection-based methods, such as A-GEM and OGD, which lead to task-agnostic versions of them. Experiments on real datasets demonstrate the effectiveness of the proposed strategy and its promising performance compared to state-of-the-art methods.