About the Project

A data analysis project studying Citi Bike usage patterns around Columbia University

Project Team

Afroditi Fragkiadaki

Afroditi Fragkiadaki

MS in Business Analytics

GitHubLinkedIn
Andrew Suh

Andrew Suh

MS in Business Analytics

GitHubLinkedIn
Naiyapak Boondee

Naiyapak Boondee

MS in Business Analytics

GitHubLinkedIn
Zhesan Liu

Zhesan Liu

MS in Business Analytics

GitHubLinkedIn
Zicheng Ni

Zicheng Ni

MS in Business Analytics

GitHubLinkedIn

Project Information

This project was completed for IEOR 4523: Data Analytics at Columbia University (Fall 2025). The objective was to perform comprehensive data analysis on a real-world dataset and develop predictive models to extract actionable insights.

We analyzed Citi Bike usage patterns around Columbia University, examining 529,908 trips from January 2024 to October 2025 across 7 stations in the Morningside Heights and Manhattanville area. Our analysis includes temporal pattern exploration, user behavior insights, and an XGBoost machine learning model for hourly demand forecasting (R² = 0.722).

Technologies Used: Python (pandas, NumPy, plotly, scikit-learn, XGBoost), Next.js, React, FastAPI, TypeScript, TailwindCSS, deck.gl, MapLibre

Data Sources: Citi Bike System Data (publicly available historical CSV files) and GBFS API (real-time station status)