The rapid growth of urban bike-sharing systems highlights their potential as sustainable transportation solutions, yet operational challenges persist. This project analyzes data from Capital Bikes, a leading bike-sharing program in Washington, D.C., to uncover patterns and optimize operations.

Key Findings:

• 30.8% dominance of weekday rides by members and a significant weekend contribution by casual riders (25.8%).
• Peak demand aligns with commuting hours (5 PM), emphasizing the need for dynamic fleet allocation.
• Predictive Modeling Results: Random Forest was the most effective, achieving an RMSE of 2.4317, capturing complex non-linear relationships.

Future Directions:

• Incorporating traffic and demographic data.
• Real-time predictions and user behavior studies.

This project bridges descriptive analytics and predictive modeling, offering a scalable framework for improving bike-sharing programs across the United States.

Objective: To analyze daily bike rentals, station usage, and weather conditions in Washington, DC, to optimize bike station placement and inventory management.

Tools Used: R, Kaggle