Data Science Fellowship for Social Good Demos

This summer thirty-six data scientists from across the country to work on social good projects in Chicago. The Data Science for Social Good fellowship was led by Obama for America’s Chief Data Science Rayid Ghani and the University of Chicago’s Computational Institute.  The fellowship was funded by a donation from Eric and Wendy Schmidt.

During the course of the summer, the fellows formed into teams that worked on 11 different projects. The DSSG team stopped by the OpenGov Hack Night to talk about the achievements made over the summer.

CTA Bus Service Simulator

https://raw.github.com/dssg/cta-otp/master/images/OTPAnalyst.png

Screenshot courtesy of DSSG

One of the most common problems along certain Chicago Transit Service bus routes is the ‘bunching’ problem that occurs when busses run too close together. The lead bus tends to be over crowded while being followed by one or more empty busses.

In order to help the CTA better optimize the bus schedule, the CTA worked with the DSSG fellows to build a simulator for the #6 bus route. The simulator builds on the core features of OpenTripPlanner‘s routing and analyst engines.

Using the simulator, the CTA will be able to see the effect of changes to the bus schedule without having to commit real world resources.

Garbage Grid System Analysis

The DSSG team also ran an analysis on if the new grid garbage system was more efficient than the old system. DSSG staff are still writing the report and we hope it will be released soon.

Evaluating Home Visitation Programs

Another project the DSSG team worked on was evaluating home visitation programs. DSSG partnered with the Nurse – Family Partnership to find the effect of home visitation programs on children’s health.

The team  studied data on low birth weights, breast feeding, and mother’s employment. While the Nurse – Family Partnership had extensive data on their own clients, they didn’t have data on residents outside their programs. So, the DSSG team developed a national immunization survey to enable comparisons.

The Data Science for Social Good fellowship program will have full reports on each of it’s projects later this month.