Reports and Conference Presentations

The pdfs below contain data and source code created in support of the IMI Project.

Hop™ transaction data was especially challenging to incorporate as there were security and privacy concerns around the personal information attached to this dataset and the potential for data users to identify a single rider’s trip patterns from the raw data.

Consulting with the TriMet Legal department was critical to the security of the data and establishing the protocols for sharing it, such as limiting to authorized audiences. Multiple discussions resulted in properly safeguarding riders’ personal information without overly restricting the ability to use and analyze the data. As a result, only data schemas are provided in the Mobility Data section in lieu of raw transaction data to avoid sharing Personally Identifiable Information (PII). However, this data schema can be used to configure data inputs that can be used by the Origin-Destination-Transfer (ODX) model.

 

Open Payments
PDF

Improve Arrival Predictions with Machine Learning
PDF

Mobility Data
PDF

 

Select models and tools that were developed as part of the IMI Project are open source and are available online at the following repositories:

• Improve Arrival Predications with Machine Learning - transitcast Github
• Improve Arrival Predications with Machine Learning - pytransitcast Github
• Origin-Destination-Transfer (ODX) Model Github