Antti Koskela, Tejas Kulkarni, Elias Raninen · 2026 · arXiv (Cornell University)
DOI: 10.48550/arxiv.2610.31069 · cited by 0 works (OpenAlex) · open-access PDF
Consideramos a otimização diferencialmente privada (DP) na configuração federada, onde uma estimativa DP da matriz de covariância da função serve como uma aproximação hesiana para modelos lineares generalizados. Os métodos DP centrais existentes para estimativa de covariância têm custos de comunicação que crescem quadraticamente na dimensão de dados, enquanto as linhas de base existentes com custos de comunicação linear são apenas localmente privadas e degradadas à medida que o número de clientes aumenta. Como remédio, propomos um quadro DP-PCA federado que atinge custos de comunicação lineares enquanto fornece garantias de DP central sob agregação segura. Consiste em um passo de seleção de rank DP, um método de potência DP e uma liberação ruidosa da covariância projetada no subespaço de baixa dimens
Machine translation (Meta M2M100); the English original follows.
We consider differentially private (DP) optimization in the federated setting, where a DP estimate of the feature covariance matrix serves as a Hessian approximation for generalized linear models. Existing central DP methods for covariance estimation have communication cost that grows quadratically in the data dimension, while existing baselines with linear communication cost are only locally private and degrade as the number of clients increases. As a remedy, we propose a federated DP-PCA framework that achieves linear communication cost while providing central DP guarantees under secure aggregation. It consists of a DP rank selection step, a DP power method, and a noisy release of the covariance projected onto the selected low-dimensional subspace. We give error bounds for this covariance estimate and for the resulting low-communication preconditioned DP gradient descent. Empirically, our proposed DP optimization method matches a centralized baseline with quadratic communication cost at a fraction of the communication.
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
Topics: Covariance, Hessian matrix, Covariance matrix, Computer science, Mathematical optimization, Estimation of covariance matrices
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.31069, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange