11:00 AM - 12:00 PM
Seminar Series: James H. Faghmous, PhD
Machine Learning for the Triple Aim: Advances, Challenges, and Opportunities
Please register here for James H. Faghmous' talk on May 22nd, 2017.
The emergence of powerful machine learning methods has captured the imagination of scientists across disciplines, including those seeking to achieve more efficient and equitable healthcare through advances in population health science and clinical research. This excitement is highlighted by the Precision Medicine Initiative that seeks to use sophisticated methods to combine complex clinical and genetic information for individualized prognosis and treatment. However, the development of machine learning efforts has been inexorably tied to the needs of the Internet industry (not population health or scientific inquiry). In this talk, I will introduce a population health audience to the nuances and limitations of machine learning, with the goal of advancing its appropriate use in the context of the health sciences. I will present joint work that highlights innovative machine learning applications and how they differ from the tools commonly available to health and policy researchers. Finally, I will also describe why (and how) we must thoughtfully introduce these technologies in a way that puts users, including physicians and population health researchers, first in order to unlock the potential of health data. Joint work with: Sanjay Basu, Aaron Baum, Emilie Bruzeillus, Ansu Chatterjee, Patrick Doupe, Vipin Kumar, Daniel Neill, Joseph Scarpa, Blanca Villanueva and other gracious collaborators.