top of page

TEAM

Subway Science

Jack Carlisle, Nicholas Haubrich

clear.png

We built and evaluated models for predicting hourly ridership of the NYC subway system, considering both the total ridership of the system and the ridership at each station. We find that an LSTM performs best for modeling the total ridership and that a CNN approach works best for the many-station case. We incorporate novel data visualization techniques to illustrate our dataset and our model's predictions.

Screen Shot 2022-06-03 at 11.31.35 AM.png
github URL

©2017-2025 by The Erdős Institute.

bottom of page