Public Transit Prediction During COVID-19 Pandemic

Joshua Smith, Yingying Zhu, Zheng Li

Abstract: Public transit demand is an import indicator of economic and social activity level. To accurately predict the public transport demand change during the COVID pandemic, in this paper, we investigate various factors affecting such demand change and collect related data from multiple sources. Different prediction models including linear regression and deep neural networks are explored. Experiments were conducted and the results show that though COVID-19 pandemic greatly affect the public transport, our proposed approach can accurately predict the next day public transit volume.