PUBPOL 5391

Unstructured Data Science Modeling

Course Description

Course information provided by the 2026-2027 Catalog.

This graduate-level course is designed for students interested in learning the foundations of unstructured data analytics. The focus will be on applying these techniques to applications in specific policy related scenarios. We will cover the intuition of the theoretical underpinnings, but the focus here is on using tools to understand and use unstructured data. We will cover topics manifold learning, clustering, topic modeling, and deep neural networks. We will use Python and learn to utilize Jupyter Notebooks. Using these tools, students will learn the underpinnings of each method so that they can choose the most appropriate for particular applications. This course is designed for students in the Jeb E. Brooks School of Public Policy MS in Data Science for Public Policy program. Other students may only enroll with permission of the instructor. As a graduate-level course, students are expected to have thoroughly read all materials prior to class and be well-prepared to discuss readings and cases with colleagues.

Prerequisites PUBPOL 5390.

Enrollment Information Enrollment limited to: graduate and professional students.

Last 4 Terms Offered (None)

Learning Outcomes

• Identify possible structure within unstructured data using visual and other techniques.
• Compare and contrast the strengths and weaknesses of various unstructured modelling techniques.
• Use Python to do exploratory and prediction with real world unstructured data.
• Describe good data and modelling practices.

View Enrollment Information