提出新方法缓解垂直联邦学习中的知识遗忘问题。
Vertical Federated Continual Learning via Evolving Prototype Knowledge
- 通过原型动态演化实现跨任务知识迁移。
- 在持续学习中提升模型性能，较先进方法提升超35%。
- 适合需要长期学习且保护数据隐私的场景。
垂直联邦学习（VFL）作为一种保护隐私的机器学习框架，用于样本对齐的特征联合。然而，传统VFL方法未解决类别和特征的持续学习挑战，导致先前任务知识出现灾难性遗忘。为此，我们提出一种新型垂直联邦持续学习方法——基于演进原型知识的垂直联邦持续学习（V-LETO），主要通过原型演化实现先前任务知识的迁移。具体地，我们设计了原型演化机制，使全局模型同时保留以往与当前任务的知识；此外，引入模型优化技术，通过限制本地模型特定参数的更新，缓解旧知识遗忘，从而提升整体性能。大量实验在类别持续学习（CIL）和特征持续学习（FIL）设置下表明，本方法优于现有最先进方法，分别在CIL和FIL任务上提升10.39%和35.15%。代码已公开于 https://anonymous.4open.science/r/V-LETO-0108/README.md。
原文摘要 · Abstract (English)
Vertical Federated Learning (VFL) has garnered significant attention as a privacy-preserving machine learning framework for sample-aligned feature federation. However, traditional VFL approaches do not address the challenges of class and feature continual learning, resulting in catastrophic forgetting of knowledge from previous tasks. To address the above challenge, we propose a novel vertical federated continual learning method, named Vertical Federated Continual Learning via Evolving Prototype Knowledge (V-LETO), which primarily facilitates the transfer of knowledge from previous tasks through the evolution of prototypes. Specifically, we propose an evolving prototype knowledge method, enabling the global model to retain both previous and current task knowledge. Furthermore, we introduce a model optimization technique that mitigates the forgetting of previous task knowledge by restricting updates to specific parameters of the local model, thereby enhancing overall performance. Extensive experiments conducted in both CIL and FIL settings demonstrate that our method, V-LETO, outperforms the other state-of-the-art methods. For example, our method outperforms the state-of-the-art method by 10.39% and 35.15% for CIL and FIL tasks, respectively. Our code is available at https://anonymous.4open.science/r/V-LETO-0108/README.md.
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。