COS597C: Bayesian Nonparametrics

Syllabus

  1. Introduction to nonparametric Bayesian statistics
    (Friday September 21 1:30PM AI Lab) [scribe notes]

  2. MCMC Sampling for Dirichlet process mixtures
    (Monday September 24 1:30PM CS402) [scribe notes]

    Assigned reading:

    Additional background reading (optional):

  3. Original constructions: Ferguson, Antoniak, and Urn schemes
    (Monday October 1 1:30PM CS402) [scribe notes]

    Assigned reading:

    Optional reading:

  4. Implement Neal's algorithm 3. (no additional reading.) (Monday October 8 1:30PM CS402) [scribe notes]
  5. Sethuraman's stick-breaking construction
    (Monday October 15 1:30PM CS402)

  6. Variational inference for Dirichlet process mixtures
    (Monday October 22 1:30PM CS402) [scribe notes (Sekora)] [scribe notes (Wang)]

  7. Hierarchical Dirichlet processes and document modeling
    (Monday November 5 1:30PM CS402) [scribe notes]

  8. NP Bayes extensions to sequential modeling
    (Monday November 12 1:30PM CS402)

    Guest discussant: Emily Fox (MIT)

    Reading:

  9. The Indian Buffet Process
    (Monday November 19 1:30PM CS402) [scribe notes]

  10. NP Bayes and computer vision
    (Monday November 26 1:30PM CS402)

  11. Random Combinatorial Structures
    (Friday November 30 1:30PM CS402) [scribe notes]

  12. Dirichlet processes and consistency
    (Wednesday December 10 1:30PM AI lab)
  13. Class presentations and course summary