A0551
Title: Bayesian latent class model for multi-source domain adaptation
Authors: Zehang Li - University of California, Santa Cruz (United States) [presenting]
Abstract: Distribution shift is a major challenge to deploying statistical and machine learning algorithms in many fields. We will discuss the challenge of distribution shift across different domains and methods to mitigate the issue in the context of assigning causes of death using verbal autopsies (VA). Worldwide, two-thirds of deaths do not have a cause assigned. VA is a well-established tool to collect information describing deaths outside of hospitals by conducting surveys with caregivers of a deceased person. It is routinely implemented in many low- and middle-income countries. Statistical algorithms to assign the cause of death using VAs are typically vulnerable to the distribution shift between the data used to train the model and the target population. This presents a major challenge for analyzing VAs as labeled data are usually unavailable in the target population. A Latent Class model framework is discussed for VA data that jointly models VAs collected over multiple heterogeneous domains, assigns the cause of death for out-of-domain observations, and estimates cause-specific mortality fractions for a new domain. We introduce a parsimonious representation of the joint distribution of the collected symptoms using nested latent class models and develop an efficient algorithm for posterior inference.