Using machine-learning models to understand nonlinear relationships between land use and travel

The number of machine learning applications to explore the complex relationship between land use and travel behavior has proliferated during the past several years. This discussion paper underscores the advantages and limitations of using machine learning approaches to reveal the complexity, compared to conventional parametric statistical models. We highlight the new insight derived from recent machine learning applications and call for further research to scrutinize the complex relationship and, especially, uncover the threshold association between land use and travel behavior.