Beyond the norm: Identifying rare and high-risk pedestrian crash patterns using unsupervised learning

Key Points

• High-risk pedestrian crash patterns were identified using unsupervised learning methods, improving safety assessments.
• Key metrics reveal emerging patterns across various urban environments, with implications for public policy and urban planning.
• Analysis utilized unsupervised learning techniques to dissect complex data sets related to pedestrian safety.
• These findings highlight the need for targeted interventions and ongoing studies to ensure pedestrian safety in cities.