Applying Differential Privacy to Tensor Completion

TL;DR

This paper introduces a unified framework for applying differential privacy to tensor completion methods, specifically CP and Tucker decompositions, balancing privacy guarantees with high accuracy in filling missing tensor data.

Contribution

It provides the first comprehensive approach to incorporate differential privacy into tensor completion, with rigorous privacy guarantees and empirical validation.

Findings

Achieves high accuracy in tensor completion under differential privacy.

Provides rigorous privacy guarantees for CP and Tucker methods.

Demonstrates effectiveness on synthetic and real datasets.

Abstract

Tensor completion aims at filling the missing or unobserved entries based on partially observed tensors. However, utilization of the observed tensors often raises serious privacy concerns in many practical scenarios. To address this issue, we propose a solid and unified framework that contains several approaches for applying differential privacy to the two most widely used tensor decomposition methods: i) CANDECOMP/PARAFAC~(CP) and ii) Tucker decompositions. For each approach, we establish a rigorous privacy guarantee and meanwhile evaluate the privacy-accuracy trade-off. Experiments on synthetic and real-world datasets demonstrate that our proposal achieves high accuracy for tensor completion while ensuring strong privacy protections.