Dynamic Behavioral Analytics in Weight-Loss Incentive Design Based on Personal Health Data

This paper aims to design reasonable exercise goals dynamically to help achieve weight loss effectively. To provide reasonable guidance for daily exercise, the personal health data based behavioral analytics framework is proposed. Then, a regulatory mechanism is utilized to avoid the abnormal effect of unreasonable goals. The parameters of the utility function, which describes the decision-making process, are estimated by the inverse optimization method. The impacts of bounds of daily steps are simulated. The results show that weight loss cannot be achieved when the step goal does not match personal ability. In addition, the designed adaptive regulatory mechanism for the step goal has a positive effect on the promotion of daily exercise.