A Comparative Study of Deep Learning Architectures for Multi-Horizon Behavioural Forecasting for Mobile Health
arXiv:2606.14604v1 Announce Type: cross Abstract: Wearable devices and smartphones generate rich behavioural time series that can support proactive health interventions, yet systematic comparisons of modern forecasting architectures for these data are lacking. In particular, it remains unclear how models generalise across populations, how different architectures respond to participant-level fine-tuning and how forecasting accuracy degrades across multi-day horizons. We benchmark six deep learning architectures, two zero-shot Foundation Models (FM) and statistical baselines on three public data
The proliferation of wearable devices and advancements in deep learning necessitate systematic comparisons for effective health intervention, marking a critical juncture for AI in personal health.
This research provides critical benchmarks for leveraging AI in predictive health, paving the way for more effective and proactive personalised healthcare interventions and potentially reducing healthcare burdens.
The systematic evaluation of deep learning architectures and foundation models for mobile health forecasting provides clearer guidance on model selection and fine-tuning for health applications.
- · Mobile health app developers
- · Wearable device manufacturers
- · Personalized medicine sector
- · Preventative healthcare providers
- · Traditional diagnostic methods
- · Generic public health campaigns
- · Inefficient healthcare systems
Improved accuracy in predicting health-related behaviors and conditions through advanced AI models.
Proactive health interventions become more feasible and widespread, leading to better public health outcomes and reduced healthcare costs.
The integration of AI-driven behavioural forecasting could transform healthcare into a highly personalized and preventive system, shifting focus from treatment to continuous wellness management.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI