In this paper, we propose the first secure federated χ2-test protocol Fed-χ2. To minimize both the privacy leakage and the communication cost, we recast χ2-test to the second moment estimation problem and thus can take advantage of stable projection to encode the local information in a short vector. As such encodings can be aggregated with only summation, secure aggregation can be naturally applied to hide the individual updates. We formally prove the security guarantee of Fed-χ2 that the joint distribution is hidden in a subspace with exponential possible distributions. Our evaluation results show that Fed-χ2 achieves negligible accuracy drops with small client-side computation overhead. In several real-world case studies, the performance of Fed-χ2 is comparable to the centralized χ2-test.
@article{arxiv.2105.14618,
title = {FED-$\chi^2$: Privacy Preserving Federated Correlation Test},
author = {Lun Wang and Qi Pang and Shuai Wang and Dawn Song},
journal= {arXiv preprint arXiv:2105.14618},
year = {2021}
}