Academics

Foundation courses cover topics that are typically included in a master’s degree in data science. If a student has completed previous graduate coursework or has relevant work experience in a foundational course topic, the foundational course requirement may be waived. Typically, students take foundation courses before enrolling in core courses.

Minimum 18 credits  

Fall Semester:  

6200 Computational I Fundamentals  

Introduces fundamental concepts of computation, data structures, algorithms, and databases, focusing on their role in data science. Covers both theoretical studies & hands-on learning activities. Includes basic data structures, advanced data structures, searching, sorting, greedy algorithms, linear programming, and basics of databases. Will develop computational thinking skills and learn a variety of ways to represent and analyze real-world data.

6300 Theory I Probability & Stochastic Processes

Covers the fundamentals of probability and stochastic processes. Students will become conversant in the tools of probability, clearly describing and implementing concepts related to random variables, properties of probability, distributions, expectations, moments, transformations, model fit, sampling distributions, discrete and continuous time Markov chains, and Brownian motion.

6400 Advanced Machine Learning I: Introduction  

Introduction to regression modeling. Topics will be discussed first in the context of linear regression, and then revisited in the context of logistic regression, ordinal regression, proportional hazards regression, and random forests. Students will be required to fit the models (both MLE and Bayesian) and use the strategies discussed in class.

6600 Data Engineering I: Data Management & Visualization  

Covers data pipeline: techniques to collect data, organize, query & apply the data, and generate products that describe the insights. Topics include Python environments, containers using Docker, data wrangling with pandas, data acquisition via flat files, APIs, JSON formats, and webscraping, relational, document, and graph databases, exploratory data analysis including static & interactive data visualization, dashboards, and cloud computing.

Spring Semester  

DS 6210 Computation II: Numerical Analysis & Optimization  

Many problems in data science essentially boil down to some mathematical relationships that are to be solved numerically. But have you ever wondered how computers could do math? This graduate-level data science course aims to cover fundamental topics of scientific computing, specifically selected and curated for data scientists, including numerical errors, root finding algorithms, numerical linear algebra, and numerical optimization.

DS 6310 Theory II: Inference and Prediction  

Explores the mathematical foundations of inferential and prediction frameworks commonly used to learn from data. Frequentist, Bayesian, Likelihood viewpoints are considered. Topics include principles of estimation, optimality, bias, variance, consistency, sampling distributions, estimating equations, information, Bootstrap methods, ROC curves, shrinkage, and some large-sample theory, prediction optimality versus estimation optimality.

DS 6410 Advanced Machine Learning II: Methods & Application  

Fundamentals of data mining and machine learning within a common statistical framework. Topics include boosting, ensembles, Support Vector Machines, model-based clustering, forecasting, neural networks, recommender systems, market basket analysis, and network centrality.

6700: Value I: Data Ethics, Policy and Governance  

Combines topics in data ethics, critical data studies, public policy, governance, and regulation. Address challenges by topic (Health, Education, Culture & Entertainment, Security & Defense, Cities, Environment, Labor). Research how data-centric systems are deployed within socioeconomic ecosystems and shape the world. Interrogate connections between data science, governments, industry, civil society organizations, and communities. 

Students must pass each core course with a minimum grade of B-.

*Will accept equivalent coursework only if course is not offered in a timely manner.

Core courses cannot be waived or substituted  

15 credit hours

Fall Semester  

DS 7200: Computation III: Distributed Computing  

Learning tools and concepts for computing on big data. Learn how to use Spark for large-scale analytics and machine learning. Spark is an open-source, general-purpose computing framework that is scalable and blazingly fast. Fundamental data types and concepts will be covered (e.g., resilient distributed datasets, DataFrames) along with Tools for data processing, storage, and retrieval, including Amazon Web Services (AWS).

DS 7400: Advanced Machine Learning III; Deep Learning  

Covers advanced theoretical concepts for deep neural networks. Topics include convolutional neural networks and their design principles, encoder-decoder architectures, recurrent neural networks, transformers, bounding box detection, image segmentation, generative adversarial networks, diffusion models, etc. Using open-source Python libraries such as NumPy, TensorFlow, and Keras, to understand how theoretical concepts are implemented.

DS 7700: Value II: Data and Society  

Introduces ways that data and information have historically been constructed in different realms — from medicine to public health to computing — to shed light on the power relationships embedded in some of our present-day and near-future tools, systems, and economic relationships. Will use a historical lens, as well as methods from STS, to introduce how data and power interact in people’s lives.

DS 7800: Research Methods in Data Science  

Transition into principal investigators and generators of data science-based knowledge. Develop practical skills necessary to conduct high-quality data science research, advance development into producers and critical consumers of research, and further development into professional data scientists broadly defined. Research based career topics covered: time management, research products, types of research positions, and grant writing.

Spring Semester  

DS 7600: Design I: Human-Centered Design  

This graduate-level course explores the principles, methodologies, and applications of human-centered design (HCD). Students will learn how to create solutions that meet real human needs. The course combines theoretical foundations with practical application through hands-on projects where students will define problems, prototype solutions, and test with users to develop impactful, user-centered products and services.

DS 7810: How to Fail at Data Science  

Elective coursework must be approved by the student’s research advisor. Students may request and receive approval to complete electives from elsewhere in the university, to gain specific knowledge or skills necessary for their research.

Electives are offered on a rotating schedule.

Refer to the Academic record for a list of electives:  Program: Data Science, Ph.D. - University of Virginia

6 credits hours  

The comprehensive exam is a multi-part, in-person exam that draws on material covered in the foundation courses. Each part is graded individually, with three possible outcomes: PhD pass, MS pass, or fail. Students must pass all parts; at least two parts must receive a PhD pass. The exam is administered by the school once a year. Students who do not pass may retake the exam once. The exam is required even if the foundation courses are waived. 

DS 7800: Data Science Research Methodology

Transition into principal investigators and generators of data science-based knowledge. Develop practical skills necessary to conduct high-quality data science research, advance development into producers and critical consumers of research, and further development into professional data scientists broadly defined. Research based career topics covered: time management, research products, types of research positions, and grant writing. 

3 credit hours  

Students typically spend the first summer in the program working in one or more faculty research labs, experiencing different research topics and environments.  

DS 8998: Master’s Level Thesis Research  

Engages students in identifying a research question, a review of the literature, and the application of an existing data science tool or technique (algorithm) to that problem. This is a mentored experience and will allow the student to demonstrate their capacity for research and begin to develop a relationship with a faculty mentor in data science. The course requires instructor permission. 

Up to 9 credit hours  

The qualifying exam is both a written and oral exam to assess the research readiness of Ph.D. candidates. The exam is administered by a qualifying committee of three faculty members, including the student’s faculty advisor. The exam centers on topics and projects proposed by the student, then vetted and approved by the qualifying committee.  If the first attempt is not successful, students may retake the exam once.  Unless an exemption is granted by the program director, the exam must be completed within one year of completing the core courses.

Successful completion of the qualifying exam marks the start of the research phase. The student will form a dissertation committee of four faculty members, including a research advisor. After crafting a research proposal, the student will publicly present the plan to the committee. Students should aim to complete the dissertation proposal within one year of completing the qualifying exam. Following the oral proposal, the examiners will decide if the student passed, conditionally passed, or failed the exam.  Students who fail may make one more attempt.  

During the research phase, the student will meet regularly with the research advisor and twice yearly with the dissertation committee. Upon successful execution of the dissertation proposal and authorship of the dissertation document, the student will present the research to the dissertation committee and the UVA community. The Ph.D. in data science is a research focused degree. Students are expected to generate new knowledge and push the boundaries of data science in their domain of choice, as well as demonstrate the impact of, and need for, these ideas in comprehensive application. Following the oral exam, the committee will decide if the student passed,