Master’s of Data Science

In this course, students will develop their programming skills while learning the fundamentals of data structures and algorithms. Students will hone their programming skills by completing non-trivial programming assignments in Python, and they will be introduced to important programming methodologies and skills, including testing and debugging. Students will learn a variety of data structures, from the basics, such as stacks, queues, and hash tables, to more sophisticated data structures such as balanced trees and graphs. In terms of algorithms, the focus will be on the practical use and analysis of algorithms rather than on proof techniques.

Syllabus

Calvin Lin

Professor, Computer Science

Probability and Simulation Based Inference for Data Science is a statistics-based course necessary for developing core skills in data science and for basic understanding of regression-based modeling. Students can look forward to gaining a foundational knowledge of inference through the simulation process.

What You Will Learn

Syllabus

Mary Parker

Associate Professor, Statistics and Data Sciences

Foundations of Regression and Predictive Modeling is designed to introduce students to the fundamental concepts behind the relationships between variables, more commonly known as regression modeling. Learners will be exposed not only to a theoretical background of the regression models, but all models will be extensively demonstrated using regression. The emphasis throughout will be on hypothesis testing, model selection, goodness of fit, and prediction.

What You Will Learn

Syllabus

Stephen Walker

Professor, Mathematics & Statistics and Data Sciences

This class covers advanced topics in deep learning, ranging from optimization to computer vision, computer graphics and unsupervised feature learning, and touches on deep language models, as well as deep learning for games.

Part 1 covers the basic building blocks and intuitions behind designing, training, tuning, and monitoring of deep networks. The class covers both the theory of deep learning, as well as hands-on implementation sessions in pytorch. In the homework assignments, we will develop a vision system for a racing simulator, SuperTuxKart, from scratch.

Part 2 covers a series of application areas of deep networks in: computer vision, sequence modeling in natural language processing, deep reinforcement learning, generative modeling, and adversarial learning. In the homework assignments, we develop a vision system and racing agent for a racing simulator, SuperTuxKart, from scratch.

What You Will Learn

Syllabus

Philipp Krähenbühl

Assistant Professor, Computer Science

This course focuses on core algorithmic and statistical concepts in machine learning.

Tools from machine learning are now ubiquitous in the sciences with applications in engineering, computer vision, and biology, among others. This class introduces the fundamental mathematical models, algorithms, and statistical tools needed to perform core tasks in machine learning. Applications of these ideas are illustrated using programming examples on various data sets.

Topics include pattern recognition, PAC learning, overfitting, decision trees, classification, linear regression, logistic regression, gradient descent, feature projection, dimensionality reduction, maximum likelihood, Bayesian methods, and neural networks.

What You Will Learn

Syllabus

Adam Klivans

Professor, Computer Science

Qiang Liu

Assistant Professor, Computer Science

This course is designed to extend foundational knowledge in probability, statistical inference, and regression analysis to settings involving complex data structures. The first part of the course focuses on statistical analysis of time series and spatial data. Topics related to time-series analysis include autocorrelation, classical time-series models, state-space models, and hidden Markov models. Topics related to spatial statistics include the analysis of spatial point patterns, Gaussian processes with spatial correlation functions and prediction/kriging, and spatial autoregressive models. A primary goal is for learners to recognize settings when standard statistical methods based on assumptions of independence are inappropriate. In addition, learners will gain awareness of and experience in applying modern statistical methods for the analysis of dependent data. Applications in areas ranging from finance to ecology to public health will be emphasized.

The second part of the course focuses on inference for matrix structured data, including methods for matrix completion and denoising, clustering, network models and inference, random walks on graphs, and graph representation learning. For matrix completion, topics include the singular value decomposition, non-negative matrix factorization, and iterative optimization methods. For network models, topics include the stochastic blockmodel and its degree corrected and mixed membership variants, latent distance models, etc. Learners will be introduced to spectral methods, Bayesian methods, convex relaxation, and likelihood-based methods for inference. Learners will also gain awareness for how different sparsity regimes call for different types of regularization in spectral approaches. Random walks on graphs will be introduced as a tool for graph representation learning.

What You Will Learn

Syllabus

Catherine Calder

Professor, Chair, Statistics and Data Sciences

Purnamrita Sarkar

Assistant Professor, Statistics and Data Sciences

In Data Exploration, Visualization, and Foundations of Unsupervised Learning, students will learn how to visualize data sets and how to reason about and communicate with data visualizations. Students will also learn how to assess data quality and provenance, how to compile analyses and visualizations into reports, and how to make the reports reproducible. A substantial component of this class will be dedicated to learning how to program in R.

What You Will Learn

Syllabus

Claus O. Wilke

Professor and Chair, Integrative Biology

This course will introduce students to statistical methods used in health data science. Topics will include survival analysis, prediction, longitudinal data analysis, design and analysis of observational studies including propensity score analysis, and design and analysis of randomized studies including sample size and power calculations, intent-to-treat analysis, and noncompliance.

What You Will Learn

Syllabus

Layla Parast

Associate Professor, Statistics and Data Sciences

While much of statistics and data sciences is framed around problems of prediction, Design Principles and Causal Inference for Data-Based Decision Making will cover basic concepts of statistical methods for inferring causal relationships from data, with a perspective rooted in a potential-outcomes framework.  Issues such as randomized trials, observational studies, confounding, selection bias, and internal/external validity will be covered in the context of standard and non-typical data structures. The overall goal of the course is to train learners on how to formally frame questions of causal inference, give an overview of basic methodological tools to answer such questions and, importantly, provide a framework for interrogating the causal validity of relationships learned from data.  The target audience for this course is someone with basic statistical skills who seeks training on how to use data to characterize the consequences of well-defined actions or decisions.

What You Will Learn

Syllabus

Corwin Zigler

Associate Professor, Statistics and Data Sciences

The "Principles of Data Science" course, designed over 12 weeks, provides a comprehensive introduction to various types of questions in data science. It explores descriptive, exploratory, inferential, predictive, and causal questions, guiding students through real-world applications and methods for addressing these problems. Key topics include data visualization, exploratory data analysis, prediction modeling, and inferential techniques, with a focus on developing statistical evidence to explain phenomena. The course also emphasizes the importance of reproducibility, version control using Git and GitHub, and structured data science workflows.

Students will learn to describe and characterize different types of data science questions, understand the analytic iteration process, and choose tools for hypothesis testing. They will also develop skills in writing reproducible documents, implementing reproducible data analysis workflows, and managing distributed collaboration. The course features case studies, such as analyzing CO2 emissions (exploratory), asthma and allergen levels (inferential), and air pollution prediction (predictive), to provide hands-on experience.

By the end of the course, students will have a solid understanding of data science procedures, tools, and ethics, allowing them to apply data science techniques effectively in various contexts. The course covers the fundamentals of coding tools, R markdown, data wrangling, and visualization principles, ensuring students are equipped to tackle practical data science problems while maintaining reproducibility and ethical standards.

Course outline will be posted soon.

Roger Peng

Professor, Statistics and Data Sciences

This course provides an in-depth exploration of the technologies behind some of the most advanced deep learning models, including diffusion models and cutting-edge models in language and computer vision. Through hands-on assignments, students will re-implement smaller versions of these models, allowing them to gain practical experience and a deep understanding of how these AI technologies function.

The course is designed to address the fact that only a small group of individuals, with significant resources, will ever train large frontier models. However, most professionals will interact with advanced AI models, either directly or indirectly. This class is valuable because it equips students with the knowledge needed to understand how these AI models work, their limitations, and how to form an accurate mental model of their capabilities. By gaining this understanding, students will be better prepared to work with and critically assess advanced AI technologies.

By the end of the course, students will be able to:

• Comprehend the inner workings of the most advanced deep learning models.
• Train and fine-tune some of these models, providing them with practical skills and hands-on experience in deep learning.

Advances in Deep Learning offers a balance between theory and application, ensuring that students leave with a robust understanding of the core technologies driving modern AI and practical experience in working with these models.

Course outline will be posted soon.

Philipp Krähenbühl

Assistant Professor, Computer Science

This course focuses on modern natural language processing using statistical methods and deep learning. Problems addressed include syntactic and semantic analysis of text as well as applications such as sentiment analysis, question answering, and machine translation. Machine learning concepts covered include binary and multiclass classification, sequence tagging, feedforward, recurrent, and self-attentive neural networks, and pre-training / transfer learning.

What You Will Learn

Syllabus

Greg Durrett

Assistant Professor, Computer Science

This class covers linear programming and convex optimization. These are fundamental conceptual and algorithmic building blocks for applications across science and engineering. Indeed any time a problem can be cast as one of maximizing / minimizing and objective subject to constraints, the next step is to use a method from linear or convex optimization. Covered topics include formulation and geometry of LPs, duality and min-max, primal and dual algorithms for solving LPs, Second-order cone programming (SOCP) and semidefinite programming (SDP), unconstrained convex optimization and its algorithms: gradient descent and the newton method, constrained convex optimization, duality, variants of gradient descent (stochastic, subgradient etc.) and their rates of convergence, momentum methods.

Syllabus

Sujay Sanghavi

Associate Professor, Electrical and Computer Engineering

Constantine Caramanis

Professor, Electrical & Computer Engineering

This course introduces the theory and practice of modern reinforcement learning. Reinforcement learning problems involve learning what to do—how to map situations to actions—so as to maximize a numerical reward signal. The course will cover model-free and model-based reinforcement learning methods, especially those based on temporal difference learning and policy gradient algorithms. Introduces the theory and practice of modern reinforcement learning. Reinforcement learning problems involve learning what to do—how to map situations to actions—so as to maximize a numerical reward signal. The course will cover model-free and model-based reinforcement learning methods, especially those based on temporal difference learning and policy gradient algorithms. It covers the essentials of reinforcement learning (RL) theory and how to apply it to real-world sequential decision problems. Reinforcement learning is an essential part of fields ranging from modern robotics to game-playing (e.g. Poker, Go, and Starcraft). The material covered in this class will provide an understanding of the core fundamentals of reinforcement learning, preparing students to apply it to problems of their choosing, as well as allowing them to understand modern RL research. Professors Peter Stone and Scott Niekum are active reinforcement learning researchers and bring their expertise and excitement for RL to the class.

What You Will Learn

Syllabus

Peter Stone Instructor of Record

Professor, Computer Science

Scott Niekum

Adjunct Assistant Professor, Computer Science