Data Science for Social Impact

Catalog description

You have some experience coding in R or Python. You've taken a class or two in basic stats or data science. But what's next? How can you use data science skills to make the world a better place? If you're asking those questions, then "Data Science for Social Impact" is for you. In this class, you'll work in four areas where data are being used to make the world better: health care, education, detecting discrimination, and clean energy technologies. You'll work with data from hospitals, schools, police departments, and electric utilities. You'll apply causal inference, prediction, and optimization techniques to help businesses, governments, and other organizations make better decisions. You'll see the challenges that arise when analyzing real data (for example, when some data are missing, or when the randomized experiment gets implemented wrong). You'll get ideas for an impactful and meaningful senior thesis, summer internship, and future career. Concretely, you'll have weekly problem sets involving data analysis in R or python. You'll learn and apply techniques like fixed effects regression, difference-in-differences, instrumental variables, regularized regression, random forests, causal forests, and optimization. Class sessions will feature active learning, discussions, and small-group case studies. You should only enroll if you expect to attend regularly and complete the problem sets on time. Prerequisites (recommended): Experience programming in R or python, or willingness to learn very quickly on your own. A basic statistics or data science course, such as any of the following: DATASCI 112, ECON 102 or 108, CS 129, EARTHSYS 140, HUMBIO 88, POLISCI 150A, STATS 60, SOC 180B, or MS&E 125.