挑战关于灾难性遗忘的常见假设
Challenging Common Assumptions about Catastrophic Forgetting
AI总结：
本文挑战了灾难性遗忘的常见假设，提出SCoLe框架研究DNN在长任务序列中的知识积累，发现SGD训练下CF影响有限，并提出提升知识积累的简单策略。
摘要：
构建能够逐步学习并积累知识的学习智能体是持续学习（CL）研究领域的核心目标。遗憾的是，在新数据上训练模型通常会损害在过往数据上的性能。在CL文献中，这种效应被称为灾难性遗忘（CF）。CF已被广泛研究，并提出了大量方法以在非重叠任务的短序列上解决该问题。在此类设置中，CF总是导致过往任务性能出现快速且显著的下降。然而，尽管存在CF，近期研究表明，在CL回归设置中，对线性模型进行SGD训练能够积累知识。当任务重复出现时，这种现象尤为明显。我们可能会问，使用SGD或任何标准基于梯度的优化训练的DNN是否也会以这种方式积累知识。此类现象对将DNN应用于真实持续场景具有有趣的推论。事实上，标准基于梯度的优化方法在计算成本上显著低于现有的CL算法。本文中，我们研究了在使用基于梯度的算法训练的DNN中，面对数据重复出现的长任务序列时的渐进式知识积累（KA）。我们提出了一个新框架SCoLe（Scaling Continual Learning）来研究KA，并发现灾难性遗忘对使用SGD训练的DNN影响有限。当在数据稀疏重复出现的长序列上训练时，总体准确率有所提升，鉴于CF现象，这可能违背直觉。我们从经验上研究了不同数据出现频率下DNN中的KA，并提出了简单且可扩展的策略以增加DNN中的知识积累。
英文摘要：
Building learning agents that can progressively learn and accumulate knowledge is the core goal of the continual learning (CL) research field. Unfortunately, training a model on new data usually compromises the performance on past data. In the CL literature, this effect is referred to as catastrophic forgetting (CF). CF has been largely studied, and a plethora of methods have been proposed to address it on short sequences of non-overlapping tasks. In such setups, CF always leads to a quick and significant drop in performance in past tasks. Nevertheless, despite CF, recent work showed that SGD training on linear models accumulates knowledge in a CL regression setup. This phenomenon becomes especially visible when tasks reoccur. We might then wonder if DNNs trained with SGD or any standard gradient-based optimization accumulate knowledge in such a way. Such phenomena would have interesting consequences for applying DNNs to real continual scenarios. Indeed, standard gradient-based optimization methods are significantly less computationally expensive than existing CL algorithms. In this paper, we study the progressive knowledge accumulation (KA) in DNNs trained with gradient-based algorithms in long sequences of tasks with data re-occurrence. We propose a new framework, SCoLe (Scaling Continual Learning), to investigate KA and discover that catastrophic forgetting has a limited effect on DNNs trained with SGD. When trained on long sequences with data sparsely re-occurring, the overall accuracy improves, which might be counter-intuitive given the CF phenomenon. We empirically investigate KA in DNNs under various data occurrence frequencies and propose simple and scalable strategies to increase knowledge accumulation in DNNs.
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。
