A-Optimal Active Learning

TL;DR

This paper introduces an active learning method based on A-optimal experimental design, which efficiently labels data and trains deep networks by leveraging Bayesian and frequentist assumptions.

Contribution

It proposes two novel approaches for active learning using A-optimal design, one Bayesian with graph Laplacian and one frequentist, improving label estimation efficiency.

Findings

Efficient label estimation for deep network training.

Two approaches outperform traditional active learning methods.

Demonstrated high efficiency in experimental results.

Abstract

In this work we discuss the problem of active learning. We present an approach that is based on A-optimal experimental design of ill-posed problems and show how one can optimally label a data set by partially probing it, and use it to train a deep network. We present two approaches that make different assumptions on the data set. The first is based on a Bayesian interpretation of the semi-supervised learning problem with the graph Laplacian that is used for the prior distribution and the second is based on a frequentist approach, that updates the estimation of the bias term based on the recovery of the labels. We demonstrate that this approach can be highly efficient for estimating labels and training a deep network.