Outsourced Privacy-Preserving Feature Selection Based on Fully Homomorphic Encryption
Koki Wakiyama, Tomohiro I, Hiroshi Sakamoto
TL;DR
This work tackles privacy-preserving feature selection when data owners and analysts are untrusted by introducing a fully outsourced computation framework based on fully homomorphic encryption (FHE). By leveraging TFHE and sorting-network preprocessing, the authors implement a consistency-based feature selection (CWC) method that operates on encrypted data with a single interaction round, improving theoretical complexity from $O(kn^2)$ to $O(kn \log^3 n)$ time and from $O(kn^2)$ space to $O(kn)$ space. The paper presents a naive pCWC and a significantly faster improved pCWC, with the latter achieving $O(kn \log^3 n)$ time and $O(kn)$ space through prefix/suffix labeling and careful synchronization during sorting. Experimental results validate the efficiency gains on small datasets, and the approach offers strong privacy guarantees under IND-CPA via FHE, making privacy-preserving feature selection feasible in untrusted environments. Future work includes extending to non-binary data, real-valued features, and multi-party or multi-key FHE settings to broaden applicability.
Abstract
Feature selection is a technique that extracts a meaningful subset from a set of features in training data. When the training data is large-scale, appropriate feature selection enables the removal of redundant features, which can improve generalization performance, accelerate the training process, and enhance the interpretability of the model. This study proposes a privacy-preserving computation model for feature selection. Generally, when the data owner and analyst are the same, there is no need to conceal the private information. However, when they are different parties or when multiple owners exist, an appropriate privacy-preserving framework is required. Although various private feature selection algorithms, they all require two or more computing parties and do not guarantee security in environments where no external party can be fully trusted. To address this issue, we propose the first outsourcing algorithm for feature selection using fully homomorphic encryption. Compared to a prior two-party algorithm, our result improves the time and space complexity O(kn^2) to O(kn log^3 n) and O(kn), where k and n denote the number of features and data samples, respectively. We also implemented the proposed algorithm and conducted comparative experiments with the naive one. The experimental result shows the efficiency of our method even with small datasets.