Modeling Individuals' Largest Daily Trip Displacement using Extreme Value Theory

As a significant factor in urban planning and traffic forecasting, human mobility draws intensive attentions in recent years with the emergence of geo-related data. In this paper, we build models for human mobility with regards to individuals’ largest daily displacement using Extreme Value Theory (EVT) based on nearly 40 million individuals’ trajectories. It is found that the largest daily displacement can be well fitted by the Generalized Pareto Distribution (GPD) and is not heavy tailed compared to the exponential distribution. Besides, we also explored differences in distribution of the largest daily displacement according to gender and weekday. The analysis results indicate that male tend to have a larger displacement and individuals tend to travel longer on weekends.