pith. sign in

arxiv: 1209.3694 · v1 · pith:UBYRWAAVnew · submitted 2012-09-17 · 💻 cs.LG · cs.AI· cs.DS

Submodularity in Batch Active Learning and Survey Problems on Gaussian Random Fields

classification 💻 cs.LG cs.AIcs.DS
keywords activebatchgaussianproblemssurveyv-optimalitycovariancecriterion
0
0 comments X
read the original abstract

Many real-world datasets can be represented in the form of a graph whose edge weights designate similarities between instances. A discrete Gaussian random field (GRF) model is a finite-dimensional Gaussian process (GP) whose prior covariance is the inverse of a graph Laplacian. Minimizing the trace of the predictive covariance Sigma (V-optimality) on GRFs has proven successful in batch active learning classification problems with budget constraints. However, its worst-case bound has been missing. We show that the V-optimality on GRFs as a function of the batch query set is submodular and hence its greedy selection algorithm guarantees an (1-1/e) approximation ratio. Moreover, GRF models have the absence-of-suppressor (AofS) condition. For active survey problems, we propose a similar survey criterion which minimizes 1'(Sigma)1. In practice, V-optimality criterion performs better than GPs with mutual information gain criteria and allows nonuniform costs for different nodes.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.