Using Machine Learning to Better Understand and Manage Disease

Friday, June 8, 2018 -
1:00pm to 2:00pm
Room 325/326, Pyle Center, 702 Langdon St.

Speaker Name: 

Mark Craven

Speaker Institution: 

Biostatistics & Medical Informatics; Computer Sciences, UW-Madison




Supervised machine learning methods infer models that map a set of input variables to an output variable of interest when given data sets consisting of such input-output pairs. The learned models can play important roles in lending insight into a problem domain and in making informative predictions. I will discuss several applications in which my group has developed and applied machine-learning approaches to help understand and manage disease processes. These applications include (i) uncovering viral genotype determinants of host disease phenotypes, (ii) extracting databases of host-virus interactions from the scientific literature, and (iii) predicting asthma exacerbations from patients’ electronic health records.