Course Topics & Schedule

Lecture slides for the various parts of the tutorial are linked below:

  1. Nonparametric Modeling, Learning, and Inference
  2. Hierarchical Models
  3. Infinite Hidden Markov Models
  4. Infinite Hidden Markov Trees
  5. Spatial Modeling via Gaussian Processes

Specific topics covered during this half-day tutorial include:

  • Nonparametric clustering and Chinese restaurant processes
  • Nonparametric latent feature models and Indian buffet processes
  • Underlying stochastic processes: Dirichlet processes, beta processes, & stick-breaking
  • Markov chain Monte Carlo learning algorithms, including Gibbs samplers
  • Variational learning algorithms, including mean field methods
  • Infinite hidden Markov models and hidden Markov trees
  • Hierarchical clustering and topic models
  • Spatially dependent modeling via kernels and Gaussian processes
  • Illustrative applications: clustering, image denoising, image segmentation, object modeling and recognition, temporal activity and motion modeling

More information about these topics, including references, can be found on the webpage for a Fall 2011 graduate seminar, Brown CSCI 2950-P: Applied Bayesian Nonparametrics.