The Coleridge Initiative's Applied Data Analytics training program focused on Local Economic Development and Job Creation in Kansas City, Missouri.
The program was generously sponsored and hosted by the Ewing Marion Kauffman Foundation.
The Coleridge Initiative Applied Data Analytics Training Program is designed to equip public policy professionals with advanced computer science and data science skills. The program provides a hands-on introduction to data analytics topics ranging from basic SQL and Python coding to running and interpreting machine learning models. Application to real-world issues is key to the organization of the program – the present Training Program in Kansas City, MO explored local economic development and job creation.
During the program, participants were grouped into teams in which they scoped and executed a project related to Economic Development and/or Job Creation. Two template projects were prepared in parallel and can be found in the Sample Projects folder.
- The project
mo_dashboard
creates an interactive dashboard to track economic development metrics at county level across the state of Missouri and on the border with Illinois. The visualized metrics include typical employment statistics such as total jobs and average wages, as well as the more advanced QWI metrics. - The project
predict_business_vitality
builds a machine learning model to predict which employers survive in the following years based on firm characteristics, industrial sector, and geography. Potential applications include intervention on failing firms and identification of weakening industries.
Participant teams usually worked on projects similar to the ones shared here – restricting to a cohort of interest or improving the models with additional features. All teams presented their results on June 28th and 29th, 2018.
For both class notebooks and team projects, participants of the Applied Data Analytics Training Program have access to several datasets, including confidential micro-data hosted on the secure Administrative Data Research Facility platform. The datasets available for the Training Program in Kansas City, MO, include:
- Missouri State LEHD Wage and Employers Data
- Kansas City, MO, Consumer Water Data
- Kansas City, MO, Business Licenses
- Census LEHD Origin-Destination Employment Statistics (LODES)
- Kansas City Star articles from the Business section
- Illinois Department of Employment Services (IDES) data
- Illinois Department of Corrections (IDOC) data
Day 1: Computing Environment, Datasets and Projects
- Variables Notebook
- Application: Generating QWI Statistics
Day 2: Databases and Visualization
Day 3: Record Linkage
Day 4: Text Analysis and Network Analysis
Day 5: Introduction to Machine Learning
Day 6: Machine Learning for Prediction
Day 7: Inference
Day 8: Machine Learning in Practice
Day 9: Privacy, Confidentiality, and Ethics
Day 10: Class Recap and Project Work
Final Presentations