A0654
Title: Bayesian clustering of complex data
Authors: Stephen Coleman - Oregon Health & Science University (United States) [presenting]
Abstract: Clustering remains a frustratingly hard problem, and often, the tools available do not meet the requirements of a specific analysis. This can be due to a multitude of reasons, but some common examples include accounting for sources of technical variation, such as is common in data generated across multiple batches, modelling shared signals across multiple modalities in a principled fashion, and accounting for missing observations within the data. Some of the challenges of designing appropriate models for such problems are covered through a Bayesian lens, as well as the related problems of implementing and applying them. The focus is on examples from the biomedical field and systems biology.
