Predicting Usage for Capital Bikeshare stations based upon Spatial Characteristics
summary
Uses Capital Bikeshare data plus OpenStreetMap and the Smart Location Database to compute station-by-station features: population and job densities, facility counts, walking distances, and percent households without cars. Implements a Python iPython notebook workflow that exports CSV metrics and calls Biogeme+Python to fit a simple linear regression and a multinomial logit. Finds some significant coefficients (residential density in evenings, bus/education in mornings, car ownership); models are preliminary and results are limited by server access so predictions are weak.
AI-written summary · about
chapters (7)
- 0:00Introduction and context
- 3:46Data sources overview
- 7:00Spatial data integration
- 10:53Regression model results
- 15:15Alternative modeling approach
- 18:14Gravity models in transport
- 20:50Conclusion and Q&A
related talks
49mStructural estimation methods to model user behaviour in bike-sharing systems
36mSmarter Tools for (Citi)Bike-Sharing
32mTour de Bikeshare: Setting World Records with JavaScript
06mUnderstanding User Behavior in Car Sharing Services Through The Lens of Mobility: Mixing ...
41mAnalysing and predicting inner-city parking space occupancy
28m"Where, When, Why, and How: Directi
06mD3P: Data-driven Demand Prediction for Fast Expanding Electric Vehicle Sharing Systems
30mLocation Intelligence and Spatial Data Science: from geographic coordinates to business insights