Online Algorithm for Unsupervised Sensor Selection

TL;DR

This paper introduces an online algorithm for unsupervised sensor selection that balances accuracy and cost without ground truth, leveraging the Weak Dominance property to achieve sub-linear regret.

Contribution

It formulates the unsupervised sensor selection as a stochastic partial monitoring problem and proposes an optimal algorithm with theoretical regret guarantees.

Findings

Algorithm achieves sub-linear regret in experiments.

Performance validated on synthetic and real-world datasets.

Effective in scenarios lacking ground truth annotations.

Abstract

In many security and healthcare systems, the detection and diagnosis systems use a sequence of sensors/tests. Each test outputs a prediction of the latent state and carries an inherent cost. However, the correctness of the predictions cannot be evaluated since the ground truth annotations may not be available. Our objective is to learn strategies for selecting a test that gives the best trade-off between accuracy and costs in such Unsupervised Sensor Selection (USS) problems. Clearly, learning is feasible only if ground truth can be inferred (explicitly or implicitly) from the problem structure. It is observed that this happens if the problem satisfies the 'Weak Dominance' (WD) property. We set up the USS problem as a stochastic partial monitoring problem and develop an algorithm with sub-linear regret under the WD property. We argue that our algorithm is optimal and evaluate its performance on problem instances generated from synthetic and real-world datasets.