Bradley Malin, Ph.D.
I have broad research interests in data mining, management, and trustworthy computing. I believe that the analysis of large quantities of health and molecular information has the potential to refine phenotype definitions into significantly more nuanced models, as well as novel clinical concepts, which associate with differential response to interventions. My research has demonstrated that we can make rapid progress in this direction by integrating novel computing infrastructures with statistically-driven methods to learn patterns and test predictive models. However, to maximize the potential for data science in clinical investigations, we must make data available on a broad scale without violating the rights of the people to whom it corresponds. As such, a great deal of my research focuses on the development of multidisciplinary approaches that draw upon methods from computer science, biomedical knowledge modeling, policy analysis, and economics.
Check out the people on the lab's online home.
Associate Editor, ACM Transactions on Information and System Security
Editorial Board, Journal of the American Medical Informatics Association
Editorial Board, Methods of Information in Medicine
Editorial Board, Transactions on Data Privacy
Editorial Board, Proceedings of the Very Large Data Bases Endowment (PVLDB)
Scientific Program Chair (Upcoming)
Scientific Program Committee (Upcoming)
... on cryptography for genomics in Nature Medicine (6/2014)
... on EMR privacy in Genome magazine (4/2014)
... on genomics and data security (6/2013)
... on my work with the Office for Civil Rights at the U.S. Department of Health and Human Services (12/2012)
... at the University of Pittsburgh (9/10/2014)
... at Carnegie Mellon University (10/6/2014)
... at Ryerson University (10/21/2014)
... at Vanderbilt's Personalized Medicine Seminar (12/10/2014)
"Data Re-identification: Societal Safeguards"
SecureMA: Protecting Participant Privacy in Genetic Association Meta-Analysis
... Journal of Biomedical Informatics:
Limestone: High-throughput Candidate Phenotype Generation via Tensor Factorization
... New England Journal of Medicine:
Ensuring Patient Privacy in Data Sharing for Postapproval Research