Uncertainty Quantification in the Classification of High Dimensional Data

CTMS Adventures In Theory Lectures Seminar

Andrew Stuart (CalTech)

Tuesday, April 11, 2017 -
3:15pm to 4:15pm
Location: 
128 Physics

We provide a unified framework for graph based semi-supervised learning which brings together a variety of methods which have been introduced in different communities within the mathematical sciences; the unification is through an inverse problems formulation. We study probit classification, generalize the level-set method for Bayesian inverse problems to the classification setting, and generalize the Ginzburg-Landau optimization-based classifier to a Bayesian setting; we also show that the probit and level set approaches are natural relaxations of the harmonic function approach introduced in machine learning. We introduce efficient numerical methods, suited to large data-sets, for both MCMC-based sampling as well as gradient-based MAP estimation. Through numerical experiments we study classification accuracy and uncertainty quantification for our models; these experiments showcase a suite of datasets commonly used to evaluate graph-based semi-supervised learning algorithms. Finally we study continuum limits of the problem formulations, and algorithms, arising in the infinite data limit.

Collaboration with

AL Bertozzi, X Luo (UCLA)
KC Zygalakis (Edinburgh)

Last updated: 2017/11/17 - 4:22pm