Graduate Courses

The program begins with an intensive ten-week summer bootcamp. All students must satisfactorily complete the bootcamp. The bootcamp is formally composed of two courses: CSS 201S. Introduction to Computational Social Science and CSS 202S. Computational Social Science Technical Bootcamp. 

CSS 201S. Introduction to Computational Social Science (8). Overview of causal and statistical inference, data types/structure, and modeling/analytical approaches to social science data.  Topics include models of social phenomena at different scales (cognition, behavior, learning, communication, language, game theory, markets, etc.) and analysis of different types of social data (social networks, text, GIS, timeseries, etc.) Emphasis is placed on the understanding of analytical and modeling methods, their applications, and limitations.

CSS 202S. Computational Social Science Technical Bootcamp (8).  This course provides practical experience with the technical skills underpinning computational social science. Topics may include calculus, linear algebra, probability, Unix/Linux, working in the terminal, file encoding, filesystem organization, Python, numpy, pandas, scikit, notebooks, matplotlib, code style, scraping, data storage, SQL, JSON, CSV/TSV, version control, git, spreadsheets, algorithms, basic data visualization, and machine learning basics.

The courses below are additionally required:

CSS 204. Statistical Computing and Inference from Data I (6). The first of a series of intensive courses in statistical computing to draw inferences from data.  This course covers research design, causal inference, data wrangling, visualization, probability, statistical inference, and the general linear model.

CSS 205. Statistical Computing and Inference from Data II (4). The second of a series of intensive courses in statistical computing to draw inferences from data.  This course covers the generalized linear model, resampling methods, maximum likelihood estimation, and regularization. Prerequisites: PSYC 201A, POLI 204B, CSS 204 or instructor approval

CSS 206.  Machine Learning for Social Sciences. (4). An introduction to machine learning methods with applications to social science data. This course covers the entire machine learning pipeline: feature engineering, model design, tuning, training, tuning, evaluation, and validation. Emphasis is on foundational methods in supervised and unsupervised machine learning problems.  Prerequisites:  CSS 202S, MATH 18, or instructor approval.

CSS 296. Research in Computational Social Science. (2, 4, 4).  Independent research under the supervision of individual faculty members.  

Students take this course for 2 units in fall quarter,  4 units in winter quarter, and 4 units in spring quarter.  This course will always be formally supervised by a faculty member at UC San Diego, though primary project management may occur with an internship partner organization. The three-quarter CSS 296 functions as a capstone, and provides a portfolio project for each student graduating with an M.S. in Computational Social Science.  The capstone may take two forms, depending on student interest and placement availability:

1. Internship Placement. Students will be embedded with a local company or organization in a mutually beneficial arrangement that allows the student to learn from the environment as well provide a meaningful contribution to the organization.
2. Faculty Project. Students will work with a faculty member on a project of mutual interest, entailing Computational Social Science techniques.

Each year, the Computational Social Science educational committee will circulate a list of elective courses for the coming year; the list will include at least two course options for each quarter with students choosing one elective in Fall, Winter, and Spring.

Course topics may include visual computing, computational modeling, econometrics, business, data analysis, casual interference etc.