Schedule
This is a tentative schedule for the course. The content may be adapted during the semester.
| Date | Topics |
|---|---|
| Mon, Jan 30, 2023 | Introduction: class logistics, motivation for the course |
| Wed, Feb 01, 2023 | An introductory review of probability (including concepts such as entropy and divergence) |
| Mon, Feb 06, 2023 | Maximum likelihood, basics of Bayesian statistics |
| Wed, Feb 08, 2023 | Exchangeable data model (coined by Dave Blei) |
| Mon, Feb 13, 2023 | Conjugate priors |
| Wed, Feb 15, 2023 | Conditional models: linear regression, logistic regression |
| Mon, Feb 20, 2023 | Bayesian mixture models, Gibbs sampling |
| Wed, Feb 22, 2023 | Gibbs sampling (ct’d), Metropolis Hastings |
| Mon, Feb 27, 2023 | Mixed membership models, Latent Dirichlet Allocation |
| Wed, Mar 01, 2023 | Variational inference |
| Mon, Mar 06, 2023 | Coordinate ascent variational inference |
| Wed, Mar 08, 2023 | Matrix factorization |
| Mon, Mar 13, 2023 | Spring recess, no lecture. |
| Wed, Mar 15, 2023 | Spring recess, no lecture. |
| Mon, Mar 20, 2023 | Exponential family |
| Wed, Mar 22, 2023 | Generalized linear models |
| Mon, Mar 27, 2023 | Neural networks |
| Wed, Mar 29, 2023 | Deep latent-variable models |
| Mon, Apr 03, 2023 | Variational Auto-Encoders (VAEs), Score function gradients, Reparameterization gradients |
| Wed, Apr 05, 2023 | Embedded Topic Models |
| Mon, Apr 10, 2023 | Tackling discrete latent variables: Marginalization, Gumbel Softmax |
| Wed, Apr 12, 2023 | Bayesian neural networks, uncertainty estimation |
| Mon, Apr 17, 2023 | Re: Metropolis-Hastings |
| Wed, Apr 19, 2023 | Hamiltonian Monte Carlo |
| Mon, Apr 24, 2023 | Hamiltonian Monte Carlo for deep latent-variable models |
| Wed, Apr 26, 2023 | Course recap and project Q&A |
| Mon, May 01, 2023 | Course poster session: 10:00 AM to noon ET, location: Zoom. Project presentations required. |
| Mon, May 09, 2023 | Dean’s Date, project final reports due. |
The lectures and lecture slides can be complemented with readings from the two books by Kevin Murphy: “Probabilistic Machine Learning: An Introduction” and “Probabilistic Machine Learning: Advanced Topics”. We will refer to the first book as KM1 and the second book as KM2. These books are freely available online at: https://probml.github.io/pml-book/book2.html. These readings aren’t mandatory, they only serve as references.
| Date | Further Reading |
|---|---|
| Wed, Feb 01, 2023 | KM1 sections 2.1, 2.2, and 2.3 |
| Mon, Feb 06, 2023 | KM1 section 4.2, KM2 section 3.2 |
| Mon, Feb 13, 2023 | KM1 section 4.6 |
| Wed, Feb 15, 2023 | KM1 sections 11.1, 11.2, 11.3.1, 11.3.3, 11.4.1, 11.4.2, 11.7, 10.1, 10.2, 10.5 |
| Mon, Feb 20, 2023 | KM1 section 3.5, KM2 section 12.3 except 12.3.2 |
| Wed, Feb 22, 2023 | KM2 sections 12.2, 12.3.2 |
| Mon, Mar 20, 2023 | KM1 sections 3.4.1, 3.4.2, 3.4.3 |
| Wed, Mar 22, 2023 | KM1 section 12 |