Portland Greenways and Intersection Safety

An MS in Data Science capstone examining how street infrastructure affects crash risk for cyclists and pedestrians.

This project is the culmination of my MS in Data Science program. Alongside my partner Thomas Sato, I analyzed how features such as neighborhood greenways, street lighting, traffic calming devices, and speed limits relate to the frequency and severity of crashes involving bikes and pedestrians, framed within Portland’s Vision Zero initiative to eliminate traffic deaths and serious injuries.

We worked with spatial data from 2007–2022, integrating and cleaning multiple city and state datasets to build a normalized, intersection-level database. Our workflow spanned data ingestion and transformation, schema design, exploratory spatial analysis, and predictive modeling. These steps allowed us to investigate how infrastructure, traffic exposure, and environmental conditions intersect to impact crash patterns and severity outcomes.

The completed project includes a fully documented database schema, interactive visualizations, and machine learning models that evaluate crash risk factors. The final site compiles our findings, data documentation, and full paper, offering insights into how targeted infrastructure improvements can support safer streets and advance Portland’s Vision Zero goals.

View Full Capstone Project