研究人员开发了DP-NGD，一个新颖的差分隐私自然梯度下降框架，旨在提高隐私保护机器学习的效用。与忽略损失曲率的标准DP-SGD不同，DP-NGD整合了这一信息，以实现更快的收敛速度和更高的准确性。该框架通过解耦曲率估计并采用具有动态钳制的白化空间机制，解决了隐私预算消耗和训练不稳定性等挑战。 AI
影响 这项研究可能带来更高效、更准确的隐私保护机器学习模型，特别是在数据效用至关重要的场景中。
排序理由 该集群包含两篇相同的arXiv预印本，详细介绍了一个用于差分隐私学习的新研究框架。
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
arXiv:2607.05866v1 Announce Type: cross Abstract: Under a fixed privacy budget, the utility of differentially private (DP) training is ultimately determined by its optimization efficiency. Standard first-order DP optimizers such as DP-SGD rely solely on local gradients and ignore…
Under a fixed privacy budget, the utility of differentially private (DP) training is ultimately determined by its optimization efficiency. Standard first-order DP optimizers such as DP-SGD rely solely on local gradients and ignore the underlying loss curvature. This geometric bli…
arXiv:2607.04777v1 Announce Type: new Abstract: Graph-structured data is increasingly generated and stored in decentralized environments, such as social platforms, mobile applications, and edge networks, where users maintain control over their local graph data. However, collectin…
arXiv:2511.13999v2 Announce Type: replace-cross Abstract: We study the running time, in terms of first order oracle queries, of differentially private empirical/population risk minimization of Lipschitz convex losses. We first consider the setting where the loss is non-smooth and…