Conceitos Básicos
The author explores the impact of behavioral, built environment, and socio-economic factors on bus use variability using explainable machine learning.
Resumo
The study investigates spatial and temporal variability of bus use during peak hours in Beijing. Findings reveal non-linear interactions between spatial and temporal variability and trip frequency. Different built environment features moderate travel time flexibility.
Public transport is crucial for urban functions but faces challenges due to poor service quality. Understanding travel patterns' variability aids in better transit planning. Smart card data analysis provides insights into transit dynamics.
Household surveys have traditionally captured daily mobility variations. Smart card data offers richer insights into transit use patterns. New metrics like stickiness help understand regularity in transit behavior.
Questions remain about the spatial-temporal variability of transit across urban spaces, influencing factors, and relationships with behavioral features. The study aims to address these gaps using machine learning techniques.
Estatísticas
Greater distance from urban centers (>10 km) increases spatial bus use variability.
Higher availability of bus routes links to higher spatial but lower temporal variability.
Lower and higher road density are associated with higher morning spatial bus use variability.