Dougal J. Sutherland

dsutherl@cs.cmu.edu

Biography

I'm a second-year PhD student in the Computer Science Department, working with Jeff Schneider. I've mostly worked on machine learning techniques that use groups of data rather than single feature vectors as the unit of interest. The primary applications I've been focusing on are using these techniques to help understand the results of large-scale scientific simulations, and in image classification. I'm also working on active learning techniques in low-rank matrix factorization models.

I did my undergraduate work at Swarthmore College in CS, linguistics, and math.

Papers

• Nonparametric Kernel Estimators for Image Classification (2012)
  nonparametric kernel estimation for image classification
• Support Distribution Machines (2012)
  Kernel algorithms on distributions

Software

• Nonparametric Divergence Estimation
  C++ and pure Matlab implementations of nonparametric divergence estimation.
• Support Distribution Machines
  C++ package to classify sample sets from distributions.