提出新方法让模型在不断学习中不忘旧知识，还能减少标注成本。
Active Learning for Continual Learning: Keeping the Past Alive in the Present
- 用费雪信息矩阵衡量样本价值，兼顾新旧任务学习
- 相比传统方法，平均准确率提升23.8%，遗忘率降低17.0%
- 适合标注昂贵、需持续学习的现实场景
持续学习（CL）使深度神经网络能够适应不断变化的数据分布。实际中，标注成本高时，需采用主动持续学习（ACL），通过选择最具信息量的样本子集来降低标注开销。然而，传统主动学习（AL）策略仅关注新知识学习，导致先前任务出现灾难性遗忘。因此，ACL需要一种能平衡遗忘抑制与快速学习能力的新策略。本文提出AccuACL——基于累积信息量的主动持续学习方法，通过理论分析推导出以费雪信息矩阵为样本选择标准，在保证性能的同时解决其可扩展性问题。大量实验表明，AccuACL在多种CL算法下显著优于基线，平均准确率提升23.8%，遗忘率降低17.0%。
原文摘要 · Abstract (English)
Continual learning (CL) enables deep neural networks to adapt to ever-changing data distributions. In practice, there may be scenarios where annotation is costly, leading to active continual learning (ACL), which performs active learning (AL) for the CL scenarios when reducing the labeling cost by selecting the most informative subset is preferable. However, conventional AL strategies are not suitable for ACL, as they focus solely on learning the new knowledge, leading to catastrophic forgetting of previously learned tasks. Therefore, ACL requires a new AL strategy that can balance the prevention of catastrophic forgetting and the ability to quickly learn new tasks. In this paper, we propose AccuACL, Accumulated informativeness-based Active Continual Learning, by the novel use of the Fisher information matrix as a criterion for sample selection, derived from a theoretical analysis of the Fisher-optimality preservation properties within the framework of ACL, while also addressing the scalability issue of Fisher information-based AL. Extensive experiments demonstrate that AccuACL significantly outperforms AL baselines across various CL algorithms, increasing the average accuracy and forgetting by 23.8% and 17.0%, respectively, on average.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。