Structural estimation methods to model user behaviour in bike-sharing systems
summary
Presents an economics-based structural model that predicts bike-share ridership by explicitly quantifying accessibility (walking distance to stations) and bike-availability (probability of finding a bicycle) using only aggregated station usage data. Incorporates real-time availability signals and hyperlocal demand sources, speeds up computation by converting from the time domain to a local-stockout-state domain, and reports about 50% better prediction accuracy than common methods for identifying when/where to improve operations or station networks.
AI-written summary · about
related talks
36mSmarter Tools for (Citi)Bike-Sharing
21mPredicting Usage for Capital Bikeshare stations based upon Spatial Characteristics
06mUnderstanding User Behavior in Car Sharing Services Through The Lens of Mobility: Mixing ...
06mD3P: Data-driven Demand Prediction for Fast Expanding Electric Vehicle Sharing Systems
28m"Where, When, Why, and How: Directi
31mScaling up Lyft’s Growth via Causal ML and Bandits
21m"Capacity Prediction" instead of "Capacity Planning":
32mTour de Bikeshare: Setting World Records with JavaScript