How does the station-area built environment influence Metrorail ridership? Using gradient boosting decision trees to identify non-linear thresholds

摘要: Abstract To inform the station-area planning, previous studies use direct ridership models to examine relationship between built environment around stations and transit ridership. Based on this framework, study innovatively applies gradient boosting decision trees investigate non-linear effects of variables station boarding. Using Metrorail data in Washington metropolitan area, we found that characteristics collectively contribute 34% predictive power for ridership, after controlling service factors demographics. Built show threshold We further identified their effective ranges, guiding land planning stations. This highlights roles compact mixed development, number bus stops, car ownership determining station-level