Year: 2026
Type: article
Abstract: Differential privacy (DP) is the gold standard for privacy-preserving analysis but introduces a fundamental trade-off between privacy guarantees and model performance. Selecting the optimal balance is a critical challenge, framed as a multi-objective optimization (MOO) problem of discovering the Pareto front and eliciting a decision-maker’s preference. While interactive MOO offers a solution, standard approaches—which model objectives separately and rely on simple pairwise feedback—are suboptimal for DP because they do not utilize the problem structure. In this work, we propose a method, PACE (Privacy-Accuracy Curve Elicitation), that exploits two key properties to reduce this inefficiency. First, we leverage the fact that the theoretical privacy guarantee naturally serves as a constraint: maximizing accuracy for a specific privacy level generates a solution on the Pareto front. Second, to efficiently model this curve, we theoretically derive the trade-off shape for regularized logistic regression, revealing a characteristic S-curve. This motivates us to model the Pareto front using a sigmoidal function. We empirically demonstrate its effectiveness across studied DP settings. This model allows us to replace less efficient pairwise comparisons with a richer interaction scheme where decision-makers directly select their most preferred solution from the hypothetical trade-off curve. Experiments on differentially private logistic regression and deep transfer learning across six datasets show that PACE converges to the most preferred trade-off with fewer model evaluations and interactions than baselines. The code can be found at https://github.com/YangYaohong1/PACE.
Source: Aaltodoc (Aalto University)
Institutions: University of Helsinki, University of Manchester, Aalto University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7212407225