Log2LoseAI: Reinforcement Learning to Create a Framework for Personalizing Financial Incentives
University of Utah
Summary
The purpose of this study is to determine the feasibility of providing personalized incentives for dietary self-monitoring and/or interim weight loss to people enrolled in a weight-loss program
Description
In this study, community outpatients will participate in a clinician-facilitated, group-based behavioral weight-loss program for 24 weeks. Dietary self-monitoring data (input by patients via a mobile phone dietary application) and weight data (input by patients via cellular scale) will be collected by a software platform. A reinforcement learning algorithm will use data collected during the trial to predict which participants will respond to a financial incentive. Incentives will be provided to participants predicted to respond, and they will be notified of incentives via text messaging.
Eligibility
- Age range
- 18+ years
- Sex
- All
- Healthy volunteers
- Yes
Inclusion Criteria: 1. Willing to attend virtual baseline and follow-up data collection visits 2. At least 18 years of age 3. Verified obesity as defined as a BMI ≥30 kg/m2 4. Agree to attend 13 biweekly group classes (virtual) delivered by a registered dietitian 5. Agree to review study materials between classes 6. Regular access to an unshared smart phone 7. Reliable access to internet 8. Able to speak and read English 9. Desire to lose weight 10. Able to connect to a video conference call using a smartphone, tablet or computer with a webcam and microphone 11. Ability to download and use Fi…
Interventions
- Behavioralpersonalized financial incentives
Every week, a reinforcement learning algorithm will process data on weight loss, calorie logging, and incentives earned to predict that their weight loss is positively influenced by incentives.
Location
- University of UtahSalt Lake City, Utah