Mitigating sampling bias in risk-based active learning via an EM algorithm
Abstract
Risk-based active learning is an approach to developing statistical classifiers for online decision-support. In this approach, data-label querying is guided according to the expected value of perfect information for incipient data points. For SHM applications, the value of information is evaluated with respect to a maintenance decision process, and the data-label querying corresponds to the inspection of a structure to determine its health state. Sampling bias is a known issue within active-learning paradigms; this occurs when an active learning process over- or undersamples specific regions of a feature-space, thereby resulting in a training set that is not representative of the underlying distribution. This bias ultimately degrades decision-making performance, and as a consequence, results in unnecessary costs incurred. The current paper outlines a risk-based approach to active learning that utilises a semi-supervised Gaussian mixture model. The semi-supervised approach counteracts sampling bias by incorporating pseudo-labels for unlabelled data via an EM algorithm. The approach is demonstrated on a numerical example representative of the decision processes found in SHM.
Cite
@article{arxiv.2206.12598,
title = {Mitigating sampling bias in risk-based active learning via an EM algorithm},
author = {Aidan J. Hughes and Lawrence A. Bull and Paul Gardner and Nikolaos Dervilis and Keith Worden},
journal= {arXiv preprint arXiv:2206.12598},
year = {2022}
}
Comments
10 pages, 8 figures. The Tenth European Workshop on Structural Health Monitoring (EWSHM 2022), Palermo, Italy, July 2022. arXiv admin note: substantial text overlap with arXiv:2201.02555. text overlap with arXiv:2206.11616