STOR 565 -- Machine Learning (Fall 2026)
Overview
This course provides a mathematically grounded introduction to statistical machine learning for advanced undergraduate and master’s students. It emphasizes foundational principles, representative models, and the connection between theoretical understanding and practical implementation.
By the end of the course, students should be able to analyze, implement, and critically evaluate machine learning methods—and independently approach new advances in ML and AI.
Fall 2026 course website
Open the STOR 565 course website
The course website hosts current course materials and activities. Sign in and use your course access to participate. Enrolled students should use UNC Canvas for submissions and grades.
Major course themes
- Foundations: empirical risk minimization, generalization, optimization, and regularization.
- Learning with labels: classification, regression, support vector machines, and kernels.
- Learning without labels: clustering, dimensionality reduction, representation learning, and generative models.
- Learning in the loop: online learning, preference data, and feedback-driven systems.
Materials
The course pairs mathematical analysis with Python implementation, experiments on real datasets, and a final team project. The aim is to help students explain why methods work, recognize their assumptions, and approach new ML techniques independently.
Read the course design outline · Teaching philosophy and other courses