Parsing human mobility in daily life circle with machine learning and Monte Carlo simulation

Abstract

In an era marked by the swift progression of urbanization and the relentless advancement of technology, a deep analysis of the relationship between human daily mobility patterns and life circles is crucial for understanding and optimizing urban spatial structures. This study, based on mobile phone data and building utilization data, first employs machine learning algorithms to measure the relationship between types of buildings and population mapping, identifying human mobility patterns. Then, the Monte Carlo method is used to simulate population movement trends, and finally, community detection algorithms are applied to identify human daily life circles. The research finds that the distribution of human daily life circles in space shows certain patterns. The machine learning model helps us to identify complex human behavior patterns, while the Monte Carlo simulation provides a new perspective in understanding the probabilistic characteristics of these patterns. This study demonstrates the strong potential of machine learning and Monte Carlo simulation in parsing complex urban mobility data. Our findings offer data-driven insights to urban planners, aiding in the design of more rational and livable urban spatial structures.