Data Science for Health and Wellness

Abstract: Recent years have seen a boom in claims that data and AI can be leveraged to improve our physical or mental wellbeing and provide personalized regimens and/or treatments. Making scientifically and statistically rigorous Wellness claims is however difficult. Every person has different physiology and preferences which influence their response to an intervention. Even individuals can exhibit extreme variability in their responses at different times and success metrics can be murky. That said, data and AI can be powerful tools for improving our wellbeing if properly marshalled and expectations are realistic. The goal of this talk is to explain how to accomplish this. Specifically, it will cover 1) the nature of physiological and biometric data, in particular its noise and variability 2) the role of preference and the user experience for effecting change and ensuring compliance and 3) what data science tools and AI approaches are most effective for tackling Wellness problems. Finally, we will discuss the legal, ethical and practical considerations of collecting, storing and leveraging biometric and other health data.

Bio: Rob Haslinger is the Lead Data Scientist for Bose Health Division which helps consumers and patients reach their fullest human potential by living healthier, more engaged lives. After obtaining a PhD in Physics and spending time at both the Santa Fe Institute for Complex Systems and Los Alamos’s Center for Non-Linear Studies, he spent ten years at Massachusetts General Hospital and MIT as a computational neuroscientist. There he researched statistical and machine learning methods for analyzing neural data, in particular multiple neuron recordings. Most recently he was Lead Data Scientist for The Sync Project, a startup acquired by Bose, which developed generative music to help its customers relax and fall asleep quicker.