Developed a statistical modeling pipeline that uses raw electronic health record data to evaluate a patient’s risk of becoming septic at various time-points in the future. The pipeline relies on data-adaptive ensemble methods to reduce the amount of assumptions required to perform valid statistical prediction and inference. Preliminary results indicate superior performance to existing methods.

Developed an exploratory data analysis method for high-dimensional network data. Supervised by Dr. Lea Popovic and Dr. Lisa Kakinami. The method was applied to the the Quebec Adipose and Lifestyle Investigation in Youth cohort data to better understand associations between social networks and lifestyle behaviours related to youth obesity.

Worked with biostatisticians, data curators and data access managers on various projects, including the preparation of statistical reports for the Public Health Agency of Canada, the development of an R package to facilitate analysis of proprietary data, the creation of software to verify accuracy of survey responses and the creation of a public use dataset.

Created metrics to measure the offensive consistency of hockey players in the National Hockey League under the supervision of Dr. Jonathan Guillemette.