Dougal J. Sutherland
I'm a postdoc at the Gatsby Computational Neuroscience Unit, University College London, working with Arthur Gretton.
GPG key / keybase.
My research interests include:
- Learning and testing on sets and distributions: two-sample tests, evaluating and training generative models, density estimation, distribution regression/classification/outlier detection.
- Approximate kernel embeddings, e.g. random Fourier features.
- Active learning, especially in nonstandard settings like searching for large-scale patterns.
I did my Ph.D. at Carnegie Mellon University,
working with Jeff Schneider
on machine learning.
various code on github,
my crossvalidated/stackoverflow profiles,
and my Swarthmore page for older stuff from undergrad.
Below, ** denotes equal contribution.
Also available as a .bib file,
and most of these are on
Efficient and principled score estimation.
Bayesian Distribution Regression.
Understanding the 2016 US Presidential Election using ecological inference and distribution regression with census microdata.
Journal and Low-Acceptance-Rate Conference Papers
Generative Models and Model Criticism via Optimized Maximum Mean Discrepancy.
International Conference on Learning Representations
Dynamical Mass Measurements of Contaminated Galaxy Clusters Using Machine Learning.
The Astrophysical Journal
831, 2, 135.
Linear-time Learning on Distributions with Approximate Kernel Embeddings.
AAAI Conference on Artificial Intelligence
On the Error of Random Fourier Features.
Uncertainty in Artificial Intelligence
Chapter 3 / Section 4.1 of my thesis supersedes this paper, fixing a few errors in constants and providing more results.
Active Pointillistic Pattern Search.
Artificial Intelligence and Statistics
A Machine Learning Approach for Dynamical Mass Measurements of Galaxy Clusters.
The Astrophysical Journal
803, 2, 50.
Active learning and search on low-rank matrices.
Knowledge Discovery and Data Mining
Nonparametric kernel estimators for image classification.
Computer Vision and Pattern Recognition
Scalable, Flexible, and Active Learning on Distributions.
Computer Science Department, Carnegie Mellon University
Integrating Human Knowledge into a Relational Learning System.
Computer Science Department, Swarthmore College
Technical Reports, Posters, etc.
Fixing an error in Caponnetto and de Vito (2007).
List Mode Regression for Low Count Detection.
IEEE Nuclear Science Symposium
Deep Mean Maps.
Kernels on Sample Sets via Nonparametric Divergence Estimates.
Grounding Conceptual Knowledge with Spatio-Temporal Multi-Dimensional Relational Framework Trees.