正则化持续学习中的二阶影响
Regularizing Second-Order Influences for Continual Learning
AI总结：
针对持续学习回放样本的连续选择干扰问题，本文基于影响函数识别并正则化二阶影响，提出新型选择目标及高效实现，并在多个基准上验证其优于现有方法。
摘要：
持续学习旨在在非平稳数据流上学习，而不会灾难性地遗忘先前知识。当前流行的基于回放的方法通过在一个保存已见数据的小型缓冲区上进行演练来应对这一挑战，这需要精细的样本选择策略。然而，现有选择方案通常只追求最大化当前选择的效用，忽视了连续多轮选择之间的干扰。受此启发，我们在一个基于影响函数构建的框架内剖析了顺序选择步骤之间的相互作用。我们成功识别出一类新的二阶影响，这类影响会逐渐放大回放缓冲区中的偶然偏差，并损害选择过程。为正则化这些二阶效应，本文提出了一种新颖的选择目标，该目标还与两个被广泛采用的准则具有明确联系。此外，我们给出了用于优化所提准则的高效实现。在多个持续学习基准上的实验证明，本方法相较最先进方法具有优势。代码可在 https://github.com/feifeiobama/InfluenceCL 获取。
英文摘要：
Continual learning aims to learn on non-stationary data streams without catastrophically forgetting previous knowledge. Prevalent replay-based methods address this challenge by rehearsing on a small buffer holding the seen data, for which a delicate sample selection strategy is required. However, existing selection schemes typically seek only to maximize the utility of the ongoing selection, overlooking the interference between successive rounds of selection. Motivated by this, we dissect the interaction of sequential selection steps within a framework built on influence functions. We manage to identify a new class of second-order influences that will gradually amplify incidental bias in the replay buffer and compromise the selection process. To regularize the second-order effects, a novel selection objective is proposed, which also has clear connections to two widely adopted criteria. Furthermore, we present an efficient implementation for optimizing the proposed criterion. Experiments on multiple continual learning benchmarks demonstrate the advantage of our approach over state-of-the-art methods. Code is available at https://github.com/feifeiobama/InfluenceCL.
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。
