Personalized weight loss strategies by mining activity tracker data

Wearable devices make self-monitoring easier by the users, who usually tend to increase physical activity and weight loss maintenance over time. But in terms of behavior adaptation to these goals, these devices do not provide specific features beyond monitoring the achievement of daily goals, such as a number of steps or miles walked and caloric outtake. The purpose of this study is twofold. By analyzing a large dataset of signals collected by these devices, we identify significant clusters of similar behavior patterns related to user physical activities. We then examine specific patterns of step count in the context of recommendation of habits that more likely give rise to weight loss effects. The evaluation of the effectiveness of these personalized recommendations, based on a comparative study, proves how a recommender system based on the reinforcement learning paradigm is able to guarantee better performance for this task by balancing the trade-off between long-term and short-term rewards.