On the Privacy of Federated Pipelines
Authors:
Open Access
- 11 July 2021
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1975-1979
Abstract
Federated learning (FL) is becoming an increasingly popular machine learning paradigm in application scenarios where sensitive data available at various local sites cannot be shared due to privacy protection regulations. In FL, the sensitive data never leaves the local sites and only model parameters are shared with a global aggregator. Nonetheless, it has recently been shown that, under some circumstances, the private data can be reconstructed from the model parameters, which implies that data leakage can occur in FL. In this paper, we draw attention to another risk associated with FL: Even if federated algorithms are individually privacy-preserving, combining them into pipelines is not necessarily privacy-preserving. We provide a concrete example from genome-wide association studies, where the combination of federated principal component analysis and federated linear regression allows the aggregator to retrieve sensitive patient data by solving an instance of the multidimensional subset sum problem. This supports the increasing awareness in the field that, for FL to be truly privacy-preserving, measures have to be undertaken to protect against data leakage at the aggregator.Keywords
Funding Information
- German Federal Ministry of Education and Research (BMBF) (01ZX1908A, 01ZX1910D, 031L0214A)
- Horizon 2020 Research and Innovation Programme (826078, 777111)
- VILLUM Young Investigator grant (13154)
This publication has 17 references indexed in Scilit:
- Experimental Evaluation of Subgraph Isomorphism Solvers2019
- Secure genome-wide association analysis using multiparty computationNature Biotechnology, 2018
- SAFETY: Secure gwAs in Federated Environment through a hYbrid SolutionIEEE/ACM Transactions on Computational Biology and Bioinformatics, 2018
- 10 Years of GWAS Discovery: Biology, Function, and TranslationThe American Journal of Human Genetics, 2017
- Fast Principal-Component Analysis Reveals Convergent Evolution of ADH1B in Europe and East AsiaThe American Journal of Human Genetics, 2016
- Routes for breaching and protecting genetic privacyNature Reviews Genetics, 2014
- Chapter 11: Genome-Wide Association StudiesPLOS Computational Biology, 2012
- A covariance-free iterative algorithm for distributed principal component analysis on vertically partitioned dataPattern Recognition, 2011
- Genetic Epidemiology of COPD (COPDGene) Study DesignCOPD: Journal of Chronic Obstructive Pulmonary Disease, 2010
- Principal components analysis corrects for stratification in genome-wide association studiesNature Genetics, 2006