Supporting Regularized Logistic Regression Privately and Efficiently

• Published in: PloS one Vol. 11; no. 6; p. e0156479
• Main Authors: Li, Wenfa, Liu, Hongzhe, Yang, Peng, Xie, Wei
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 06.06.2016; Public Library of Science (PLoS)
• Subjects: Analysis; Biology and Life Sciences; Collaboration; Computer and Information Sciences; Computer Communication Networks - organization & administration; Computer Communication Networks - standards; Computer Communication Networks - statistics & numerical data; Computer engineering; Computer Security; Confidentiality; Consortia; Cooperative Behavior; Cryptography; Cybersecurity; Data collection; Data processing; Data retrieval; Distributed processing; Empirical analysis; Engineering and Technology; Epidemiology; Genetic Information Nondiscrimination Act 2008-US; Genetics; Genomes; Health informatics; Humans; Information Dissemination - methods; Information management; Information services; Information sharing; Information technology; Laboratories; Learning algorithms; Linear algebra; Logistic Models; Machine learning; Machine Learning - standards; Management; Mathematical models; Medicine and Health Sciences; Methods; Models, Statistical; Network analysis; Parkinson's disease; Physical Sciences; Privacy; Quantitative genetics; Regression analysis; Regression models; Regularization; Research and Analysis Methods; Security; Smart grid; Social Sciences; Statistical analysis; Statistical models; United States; Beijing China; China
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0156479

As one of the most popular statistical and machine learning models, logistic regression with regularization has found wide adoption in biomedicine, social sciences, information technology, and so on. These domains often involve data of human subjects that are contingent upon strict privacy regulations. Concerns over data privacy make it increasingly difficult to coordinate and conduct large-scale collaborative studies, which typically rely on cross-institution data sharing and joint analysis. Our work here focuses on safeguarding regularized logistic regression, a widely-used statistical model while at the same time has not been investigated from a data security and privacy perspective. We consider a common use scenario of multi-institution collaborative studies, such as in the form of research consortia or networks as widely seen in genetics, epidemiology, social sciences, etc. To make our privacy-enhancing solution practical, we demonstrate a non-conventional and computationally efficient method leveraging distributing computing and strong cryptography to provide comprehensive protection over individual-level and summary data. Extensive empirical evaluations on several studies validate the privacy guarantee, efficiency and scalability of our proposal. We also discuss the practical implications of our solution for large-scale studies and applications from various disciplines, including genetic and biomedical studies, smart grid, network analysis, etc.