Implementing and Evaluating Machine Learning Algorithms for Bikeshare System Demand Prediction

A bikeshare (public bicycle, or bicycle-sharing) system is a service in which bicycles are made available for shared use to individuals on a short-term basis for a price or free. Bikeshare systems have increased from operating in a few European cities to expanding in the United States at an increasing pace. Many bikeshare systems allow users to borrow a bike from a station and return it at another station belonging to the same system. The goal is to encourage cycling as a mode of transportation as well as recreation. Nevertheless, the flexibility to pick up and return bicycles at any station can lead to inventory imbalances in the system. To enhance the effectiveness of the system, bikeshare operators should implement suitable methods to realign resources, guided by precise forecasts of bicycle demand. This research endeavors to develop models for Houston bikeshare system demand prediction at the station level by leveraging data on station activities. Accurate prediction of bikeshare demand has the potential to transform the way these systems are managed and integrated into urban transportation networks, leading to improved efficiency, customer satisfaction, and sustainability.

Research Area