A Data-Driven Framework for Forecasting & Business Impact
This project presents a machine learning framework to forecast demand for Divvy bike-sharing stations in Chicago. The goal is to support business decisions such as resource allocation and station management using predictive analytics.
View the full project presentation on Google Slides
- Problem Statement: Forecast short-term & Long-term demand to optimize redistribution and minimize shortages.
- Tech Stack: Python, scikit-learn, XGBoost, time series analysis, geospatial clustering
- Approach:
- Station segmentation based on usage patterns
- City-wide and station-level demand forecasting
- Model evaluation and business impact analysis
