Efficient Graph-Based Active Learning with Probit Likelihood via Gaussian Approximations
Machine Learning
2020-07-23 v1 Machine Learning
Abstract
We present a novel adaptation of active learning to graph-based semi-supervised learning (SSL) under non-Gaussian Bayesian models. We present an approximation of non-Gaussian distributions to adapt previously Gaussian-based acquisition functions to these more general cases. We develop an efficient rank-one update for applying "look-ahead" based methods as well as model retraining. We also introduce a novel "model change" acquisition function based on these approximations that further expands the available collection of active learning acquisition functions for such methods.
Cite
@article{arxiv.2007.11126,
title = {Efficient Graph-Based Active Learning with Probit Likelihood via Gaussian Approximations},
author = {Kevin Miller and Hao Li and Andrea L. Bertozzi},
journal= {arXiv preprint arXiv:2007.11126},
year = {2020}
}
Comments
Accepted in ICML Workshop on Real World Experiment Design and Active Learning 2020