Algorithms and data increasingly power our private and civic life. Companies, nonprofits, and governments have invested heavily in data mining - the bulk collection of user behavior data from web platforms to understand public opinion and to forecast trends. A lot of fashionable terms, such as artificial intelligence and big data, are being thrown around these days. The public and regulators also become increasingly wary of the dark side of algorithms ? the skepticism has culminated after the Cambridge Analytica scandal and the revelation of alleged foreign propaganda in the US through social media. This course gives a practical understanding of how data mining and algorithms work. You can obtain (1) marketable computational skills in data analytics and visualization and (2) evidence-based critical perspectives on the algorithmic society we live in. This course is offered as part of the new graduate program in Data Analytics and Computational Social Science, and would count towards the technical electives course distribution requirement of the master's degree.
This course serves as a rigorous introduction to quantitative empirical research methods, designed for doctoral students in social science and master?s students with a data analytics or computational social science focus. The material covered will include a brief introduction to the problem of causality, followed by modules on (1) measurement, (2) prediction, (3) exploratory data analysis (discovery), (4) probability (including distributions of random variables), and (5) uncertainty (including estimation theory, confidence intervals, hypothesis testing, power). Along the way, we will encounter linear regression and classification as tools of descriptive data summary, prediction and inference, and as part of a broader strategy of causal analysis. Simulations and data analysis will be conducted in the R statistical environment. This course is a required core course for the graduate certificate and the master?s degree in Data Analytics and Computational Social Science (DACSS).