Winter 2021

University of Washington

It emphasizes inference in engineering settings, utilizing the powerful language of probabilistic graphical models. This course provides a good blend of probability theory, graph theory, and computation.

This course focuses on how to address inference in complex engineering settings. While driven by applications, the course is not about these applications. Rather, it emphasizes a common foundation and conceptual framework for inference-related questions arising in various fields including, but not limited to, machine learning, signal processing, artificial intelligence, computer vision, control, and communication. We focus on inference: learning about the hidden state of the world that we care from available observations. We rely on the powerful language of probabilistic graphical models to leverage on the inherent structure of the given problem and efficiently perform inference. Graphical models build upon the beautiful marriage between probability theory and graph theory and use graphs to capture the fundamental structure of multivariate statistical models and also design efficient computation for inference tasks.

Students entering the class should be comfortable with programming and should have a pre-existing working knowledge of linear algebra (e.g., MATH 308), probability and statistics (e.g., CSE 312/STAT390), and algorithms. For a brief refresher, we recommend that you consult the statistics/probability reference materials on the Textbooks page.

No data.

There is no existing text that perfectly matches the content of CSE515. However, the following references contain useful additional details and insights. Email the instructor if you are interested in further reading materials.

- D. Koller & N. Friedman, Probabilistic Graphical Models: Principles and Techniques, MIT Press, 2009
- Steffen L. Lauritzen, Graphical Models, Oxford University Press, 1996
- Marc Mezard and Andrea Montanari, Information, Physics, and Computation, Oxford University Press, 2009
- M. Wainwright and M. Jordan, Graphical models, exponential families, and variational inference, Foundations and Trends in Machine Learning, 2008

Lecture slides and notes available at Tentative Schedule

No videos available

Homework available at Homework

Readings available at Tentative Schedule

Belief propagationCausal structure discoveryDensity evolutionDirected graphical modelsGaussian Belief PropagationGaussian graphical modelsLearning graphical modelsMarkov chain Monte Carlo (MCMC)Max-product algorithmRelations between graphical modelsSum-product algorithm on factor graphsUndirected graphical modelsVariational methods