RESEARCH INTERESTS

My interests are in harmonic analysis, wavelets, multiscale analysis in general, and in particular with applications to the analysis of graphs and data sets viewed as discrete or sampled continuous geometric structures embedded in high-dimensional spaces. I am interested in machine learning problems, mostly from the point of view of approximation and fitting of functions under random noise and random sampling.

  1. Diffusion Wavelets: a recent construction of new families of wavelets and Multi-resolution Analyses on graphs, manifolds and point clouds. Pictures, papers and presentations available.
  2. Diffusion Geometries: here are some links to the use of diffusion geometries in data analysis.
  3. Multiscale Analysis of Markov Decision Processes
  4. Harmonic Analysis and Wavelets: here I talk a bit about Harmonic Analysis and provide links to related web pages.
  5. HyperSpectral Imaging and Pathology : hyper-spectral imaging applied to pathology

Current postdocs and students

Jake Bouvrie, Guangliang Chen, Prakash Balachandran, Anna V Little.

Pointers to some future, present and recent past happenings

CTMS Workshop on Large Data Sets: Computation and Structure, Nov. 13th, Duke University, NC. The schedule will be available here.

Forum on Geometric Aspects of Machine Learning and Visual Analytics, Oct 11-12, IEEE VisWeek, Atlantic City, NJ.

SAMSI opening workshop on stochastic dynamics, part of the long program on stochastic dynamics.

Our Probability Wiki and working group at Duke. As of Spring 09, we are meeting every Tuesday at 1pm in Room 259 in the Physics building, to discuss topics of interest to the audience. It is informal, highly interdisciplinary, and fun. Papers/talks are collected on the wiki. Feel free to join us (or the mailing list).

Symposium on Manifold Learning to be held on Nov. 5-7 in Arlington, VA.

Compressive Sensing Workshop at Duke (slides and video lectures available).

Links of Interest

Programming, Code, etc...

Data Sets

Some Google searches

The material presented at this web site is partially based upon work supported by the Alfred P. Sloan Foundation and the National Science Foundation. Any opinions, findings and conclusions or recomendations expressed in this material are those of the author and do not necessarily reflect the views of the granting agencies.