Title: A Bayesian Nonparametric Model for Predicting Pregnancy Outcomes Using Longitudinal Profiles

Abstract: Across several medical fields, developing an approach for disease classification is an important challenge. The usual procedure is to fit a model for the longitudinal response in the healthy population, a different model for the longitudinal response in disease population, and then apply the Bayes' theorem to obtain disease probabilities given the responses. Unfortunately, when substantial heterogeneity exists within each population, this type of Bayes classification may perform poorly. In this paper, we develop a new approach by fitting a Bayesian nonparametric model for the joint outcome of disease status and longitudinal response, and then use the clustering induced by the Dirichlet process in our model to increase the flexibility of the method, allowing for multiple subpopulations of healthy, diseased, and possibly mixed membership. In addition, we introduce an MCMC sampling scheme that facilitates the assessment of the inference and prediction capabilities of our model. Finally, we demonstrate the method by predicting pregnancy outcomes using longitudinal profiles on the $\beta$--HCG hormone levels in a sample of Chilean women being treated with assisted reproductive therapy.