Title: Topic Models, Methods, and Medicine
Speaker: Prof. Jamie Haddock (Harvey Mudd College)
Abstract: There is currently an unprecedented demand for efficient, quantitative, and interpretable methods to study large-scale (often multi-modal) data. One key area of interest is that of topic modeling, which seeks to automatically learn latent trends or topics of complex data sets, providing practitioners a view of what is “going on” inside their data. This talk will survey several new tools for topic modeling on matrix and tensor data which allow for use of various forms of supervision and which learn hierarchical structure amongst topics. These tools are of interest across the many fields and industries producing, capturing, and analyzing big data, but are of particular interest in applications where expert supervision is available and often essential (e.g., medicine). We will describe two applications of these methods to medical data; an application to a large-scale patient survey database and an ongoing application to cardiovascular imaging data.
Prof. Jamie Haddock is an Assistant Professor in the Mathematics Department at Harvey Mudd College