Title: Scalable Bayesian nonparametric clustering and classification
Authors: Yang Ni - Texas A&M University (United States) [presenting]
Abstract: A scalable multi-step Monte Carlo algorithm is developed for inference under (possibly non-conjugate) Bayesian nonparametric models. Each step is ``embarrassingly parallel'' and can be implemented using the same Markov chain Monte Carlo sampler. The simplicity and generality of our approach makes a wide range of Bayesian nonparametric methods applicable to large datasets. Specifically, we apply product partition model with regression on covariates using novel implementation to classify and cluster patients in large electronic health records study. We find interesting clusters and superior classification performance against competing classifiers.