基于递归特征归一化的无混杂因素持续学习
Confounder-Free Continual Learning via Recursive Feature Normalization
AI总结：
针对持续学习中混杂因素导致的有偏预测和灾难性遗忘问题，提出可集成于任意深度架构的递归MDN层，利用递归最小二乘回归更新内部状态，实现跨人群公平预测。
摘要：
混杂因素是同时影响输入和目标的无关变量，导致虚假相关和有偏预测。近年来，在传统模型中处理或去除混杂因素方面取得了进展，例如元数据归一化（MDN），该方法根据研究混杂因素调整学习到的特征分布。然而，在持续学习背景下，即模型随时间从新数据中持续学习而不遗忘，学习对混杂因素不变的特征表示仍然是一个重大挑战。为了从中间特征表示中去除其影响，我们引入了递归MDN（R-MDN）层，该层可以集成到任何深度学习架构中，包括视觉变换器，并且可以应用于模型的任何阶段。R-MDN通过递归最小二乘算法执行统计回归，以维持并持续更新内部模型状态，以适应数据和混杂变量分布的变化。我们的实验表明，R-MDN在静态学习和持续学习的不同阶段中，通过减少由混杂因素随时间变化引起的灾难性遗忘，促进了跨人群组的公平预测。
英文摘要：
Confounders are extraneous variables that affect both the input and the target, resulting in spurious correlations and biased predictions. There are recent advances in dealing with or removing confounders in traditional models, such as metadata normalization (MDN), where the distribution of the learned features is adjusted based on the study confounders. However, in the context of continual learning, where a model learns continuously from new data over time without forgetting, learning feature representations that are invariant to confounders remains a significant challenge. To remove their influence from intermediate feature representations, we introduce the Recursive MDN (R-MDN) layer, which can be integrated into any deep learning architecture, including vision transformers, and at any model stage. R-MDN performs statistical regression via the recursive least squares algorithm to maintain and continually update an internal model state with respect to changing distributions of data and confounding variables. Our experiments demonstrate that R-MDN promotes equitable predictions across population groups, both within static learning and across different stages of continual learning, by reducing catastrophic forgetting caused by confounder effects changing over time.
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。
