持续学习的可学习性与算法
Learnability and Algorithm for Continual Learning
AI总结：
本文证明类增量学习（CIL）是可学习的，并基于理论提出新算法，实验验证其有效性。
摘要：
本文研究了类增量学习（CIL）这一具有挑战性的持续学习（CL）设定。CIL学习一系列任务，这些任务由不相交的概念或类别集合组成。在任何时刻，构建一个单一模型，该模型可用于对迄今为止学到的任何类别的测试实例进行预测/分类，而无需为每个测试实例提供任何任务相关信息。尽管已经提出了许多CIL技术，但它们大多是经验性的。最近研究表明，一个强大的CIL系统需要每个任务具备强大的任务内预测（WP）和强大的分布外（OOD）检测能力。然而，CIL是否实际上可学习仍不得而知。本文证明了CIL是可学习的。基于该理论，还提出了一种新的CIL算法。实验结果证明了其有效性。
英文摘要：
This paper studies the challenging continual learning (CL) setting of Class Incremental Learning (CIL). CIL learns a sequence of tasks consisting of disjoint sets of concepts or classes. At any time, a single model is built that can be applied to predict/classify test instances of any classes learned thus far without providing any task related information for each test instance. Although many techniques have been proposed for CIL, they are mostly empirical. It has been shown recently that a strong CIL system needs a strong within-task prediction (WP) and a strong out-of-distribution (OOD) detection for each task. However, it is still not known whether CIL is actually learnable. This paper shows that CIL is learnable. Based on the theory, a new CIL algorithm is also proposed. Experimental results demonstrate its effectiveness.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
