Wesley's MyFitnessPal Research Talk

Room: GDC 3.516
Contact Name: 
Wesley Tansey
May 1, 2014 4:00pm - 6:00pm

MyFitnessPal (MFP) is the world's most popular health and wellness app, with over 60M users across all mobile and desktop platforms. People generally sign up to MFP to track their diet and exercise as they work toward a target weightloss goal. This presents several important machine learning problems, such as identifying the diets people are following, modeling how diets change over time, determining the effectiveness of a given diet, and making recommendations to nudge users towards a healthier diet. I'll be presenting preliminary results on applying Bayesian hierarhcical modeling to address the first two problems, as well as an outline of how the current model can be extended to help answer the final two questions. The talk should help to give an overview of my current line of research and hopefully spark a fruitful discussion on potential future lines of research.