Q: What is the Data Science Federation?
The Data Science Federation is a collaborative effort between the Information Technology Agency, 12 (and growing!) local universities and colleges, City Departments and Elected officials to bring new ideas to City challenges using data science and predictive analytics.
Q: Who are our Partners?
We partner with 12 local universities, City Departments, Elected Officials and other relevant agencies (such as Los Angeles County, Hack for LA, and local tech businesses) to deliver data science solutions that help improve the delivery or efficiency of City services or information. A full listing on our partner universities is available online.
Q: How do we select projects?
Projects are suggested by City departments or officials and then evaluated to see whether data science tools and techniques could help and whether a Federation partner school has a matching interest and skill set.
Q: How do we scope and execute projects?
Once a topic is brainstormed, we use a project scoping template to determine what exact questions we are trying to answer, what data sources will be used, and set expectations for the project. Here is an example project scoping document for our Predicting Risk of Homelessness project. After that, we’ll shop the proposal around to our University partners to find an ideal match.
From there, we set up a shared Github repository for hosting the code and discussing issues. Think of Github as a Google drive but for code. After the project completion, we work with our partners to determine deployment and next steps if applicable.
Q: What are some example projects and successes?
Helped the Office of Finance determine which businesses to audit, leading to a projected doubling in the rate of audit return.
Built an algorithm to score every rent-stabilized building in Los Angeles with an Ellis Act risk score for the next 6 months, with 98% accuracy at predicting removals for the Housing and Community Investment Department.
Helped the Bureau of Sanitation build a realtime dashboard for a variety of programs under SANSTAT, replacing a labor intensive manual process and infrequent observations with a tool that allowed up to the hour changes in operations.
Q: What else are we up to?
Host a Data Science Working Group for internal partners to help train and collaborate.
Maintain a relationship with the Los Angeles County Data Team to collaborate on regional issues.
Help organize Data + Donuts, a speaker series for Public Sector data professionals with County, Southern California Association of Governments, others.
Support multiple grants to the National Science Foundation and others on behalf of the City and our local universities.
Speak to schools, businesses, and at events to better understand the opportunities for building the data ecosystem in Los Angeles