Undergraduate Courses
Probabilistic Modeling and Machine Learning
DSC 140A
Rigorous introduction to machine learning from a probabilistic perspective. Covers foundational paradigms for learning, modeling, and estimation. Topics include: maximum likelihood estimation, empirical risk minimization, linear and logistic regression, calibration, model selection, regularization, support vector machines, generative models, and expectation maximization.
Data Science Capstone
DSC 180AB
Two-quarter capstone sequence where students work with faculty or industry mentors to design and execute data science projects in teams. Students gain domain background through guided projects and then propose and complete independent capstone projects. Covers the complete data science lifecycle: problem assessment, data collection and cleaning, model creation, ethical considerations, system design, and results presentation. Includes methodology component on software engineering, project management, and effective communication.
Graduate Courses
Machine Learning Competitions
DSC 190/291
Advanced course where students compete to build state-of-the-art machine learning models on modern prediction tasks. Students select from competitions provided by industry partners or online platforms. Applications focus on medicine, finance, and scientific discovery. Teams develop models with instructor guidance and peer feedback. Evaluation based on competition performance, development approach, and reproducibility. Requires extensive machine learning coursework and prior experience with the Python ML ecosystem.Retry
Interpretable and Explainable Machine Learning
DSC 270
Introduction to modern methods for interpretability and explainability in machine learning. Covers algorithms for learning transparent models such as rule lists, sparse linear models, and decision trees. Explores post-hoc explainability techniques including model distillation, attribution scores, and counterfactual explanations. Examines overarching concepts like cognitive biases, model multiplicity, and trustworthiness. Students implement and apply methods through programming assignments in medicine, credit scoring, and data-driven discovery.