Predicting High-Risk Behaviors in Individuals with Profound Autism Using Sleep and Other Environmental Factors
Published:
Abstract
This presentation explores the relationship between sleep architecture and daytime behaviors in individuals with profound autism spectrum disorder. Using a novel privacy-preserving sensing approach combined with machine learning techniques, we demonstrate how disruptions in sleep patterns can predict challenging behaviors in the following 24-hour period. The talk presents a framework for early intervention that could significantly improve quality of life and reduce caregiver burden for families supporting individuals with profound autism.
Key Topics
- Novel off-body sensing technologies for sleep monitoring
- Machine learning approaches for behavioral prediction
- Privacy-preserving edge computing in vulnerable populations
- Practical intervention strategies based on predictive modeling
- Clinical implications and future directions for personalized care