A1238
Title: Latent modularity in multi-view data
Authors: Andrea Cremaschi - IE University (Spain) [presenting]
Maria De Iorio - National University of Singapore (Singapore)
Garritt Page - Brigham Young University (United States)
Ajay Jasra - The Chinese University of Hong Kong Shenzhen (China)
Abstract: The problem of clustering multi-view data, that is, information associated with individuals that form heterogeneous data sources (the views), is considered. A Bayesian model is adopted where the prior structure assumes that each individual belongs to a baseline cluster, and conditionally allows each individual in each view to potentially belong to different clusters than the baseline. This structure is called latent modularity. In each view, a specific statistical model is employed with an associated prior. Expressions for the marginal priors on the view-specific cluster labels and the associated partitions are derived, providing insights into the chosen prior structure. Using a tailored Markov chain Monte Carlo algorithm, the model is showcased in a simulation study and in a clinical case study involving multi-view measurements from the GUSTO cohort.
