Matthew J. Beal - CSE 714 Advanced Topics in Machine Learning (Spring 2006)

home

research

publications

software

teaching

contact

cv/bio

personal

to students

cse 714

home

schedule

Announcements

The seminar will be letter-graded, determined by research, project, presentation, and participation.

Our regular meeting time is Monday 3-6 (first meeting Jan 23).

A selection of possible topics that we will cover are listed below:

Topic models (LDA, LSA, IR applications)
Dirichlet processes (infinite mixture models)
Gaussian processes (smoothing, prediction, classification)
Learning with unlabeled data (bootstrapping, semi-supervised, spectral methods)
Spectral methods (image segmentation, clustering, dim. reduction)
Dimensionality reduction (PCA, FA, LLE, MFA, SNE, Isomap)
Kernel machines (theory and practice)
Conditional random fields (maxent HMMs, LingX apps)
Comp.Lingx (PCFGs, WordSenseDisambiguation)
Comp.Bio (e.g. insert/delete HMMs etc.)
Sampling (Metropolis, importance, slice, exact, MCMC etc.)

Unregistered students are welcome, email me to be added to the email distribution list and UBlearns.

Matthew J. Beal • Updated 15 January 2006 • Email mbeal [at] cse.buffalo.edu