Adaptive data selection improves wearable prediction under low baseline performance
Researchers have developed adaptive sensing strategies that selectively sample data to improve prediction performance in wearable health systems. These strategies yield significant improvements in prediction accuracy for individuals with lower baseline performance, while offering minimal gains for those with strong baselines. The findings suggest that adaptive sensing is most beneficial in underperforming scenarios, supporting tailored deployment based on individual performance levels to enhance efficiency in wearable health monitoring. AI
IMPACT Adaptive sensing strategies can improve the efficiency and accuracy of AI-driven wearable health monitoring, particularly for individuals who may not initially benefit from standard models.