Identifying Transit Deserts by Using Linear Regression and Clustering Algorithms

Abstract

As cities strive to increase sustainable transportation options, understanding and addressing transit deserts—areas where public transit is insufficient to meet residents’ needs—becomes essential. This study examines transit deserts within Chicago by integrating sociodemographic data and public transit usage patterns. Through linear regression and clustering methods, key population characteristics influencing passengers’ reliance on public transit across community areas are identified. Additionally, the analysis of Divvy bike usage data highlights disparities in bike station distribution, with most stations concentrated in central Chicago. This concentration limits transportation accessibility for outer areas, which may have latent demand for increased transit options. Our findings suggest potential high-demand areas lacking adequate service, supporting the case for a strategic redistribution of transit resources. The methodology and insights of this study extend beyond Chicago, offering a framework for identifying transit deserts in other urban centers to enhance equitable transit access and improve urban mobility infrastructure.