Temporal Point Process Modeling of Aggressive Behavior Onset in Psychiatric Inpatient Youths with Autism

2025-03-20

AI Summary

This study addresses the challenge of long-term risk prediction for aggressive behaviors—including both aggression toward others and self-injury—among hospitalized adolescents with autism spectrum disorder. We propose the first temporal point process model based on the Hawkes process for this clinical domain. Unlike existing approaches limited to ultra-short-term (1–3 minute) forecasting, our method enables quantitative prediction of event probability and expected count over time windows exceeding five minutes, while uncovering critical branching dynamics (branching factor ≈ 0.97). Innovatively adapting self-exciting point processes to clinical behavioral modeling, we integrate differentiable intensity estimation, rigorous goodness-of-fit testing, and a multi-scale evaluation framework. Validated on real-world clinical data, our approach significantly improves long-horizon predictive accuracy. It delivers interpretable, quantifiable computational support for early identification of high-risk states and timely preventive interventions.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• User Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Security and Privacy: Large-scale security measurements

Abstract

Aggressive behavior, including aggression towards others and self-injury, occurs in up to 80% of children and adolescents with autism, making it a leading cause of behavioral health referrals and a major driver of healthcare costs. Predicting when autistic youth will exhibit aggression is challenging due to their communication difficulties. Many are minimally verbal or have poor emotional insight. Recent advances in Machine Learning and wearable biosensing enable short-term aggression predictions within a limited future window (typically one to three minutes). However, existing models do not estimate aggression probability within longer future windows nor the expected number of aggression onsets over such a period. To address these limitations, we employ Temporal Point Processes (TPPs) to model the generative process of aggressive behavior onsets in inpatient youths with autism. We hypothesize that aggressive behavior onsets follow a self-exciting process driven by short-term history, making them well-suited for Hawkes Point Process modeling. We establish a benchmark and demonstrate through Goodness-of-Fit statistics and predictive metrics that TPPs perform well modeling aggressive behavior onsets in inpatient youths with autism. Additionally, we gain insights into the onset generative process, like the branching factor near criticality, and suggest TPPs may enhance future clinical decision-making and preemptive interventions.

Problem

• Modeling aggression onset timing in autistic youths
• Predicting aggression probability over extended periods
• Providing interpretable insights into aggression dynamics

Innovation

• Temporal Point Processes model aggression onset
• Self-exciting Hawkes processes capture clustered behavior
• Probabilistic forecasts support clinical intervention decisions