REDS: Rule Extraction for Discovering Scenarios

TL;DR

REDS is a novel method that significantly reduces the number of simulations needed for scenario discovery by leveraging an intermediate machine learning model to label data, improving efficiency and applicability.

Contribution

The paper introduces REDS, a new scenario discovery procedure that uses an intermediate ML model to enhance subgroup discovery with fewer simulations.

Findings

Reduces simulations by 50-75% on average.

Effective as semi-supervised subgroup discovery.

Improves scenario detection from third-party data.

Abstract

Scenario discovery is the process of finding areas of interest, known as scenarios, in data spaces resulting from simulations. For instance, one might search for conditions, i.e., inputs of the simulation model, where the system is unstable. Subgroup discovery methods are commonly used for scenario discovery. They find scenarios in the form of hyperboxes, which are easy to comprehend. Given a computational budget, results tend to get worse as the number of inputs of the simulation model and the cost of simulations increase. We propose a new procedure for scenario discovery from few simulations, dubbed REDS. A key ingredient is using an intermediate machine learning model to label data for subsequent use by conventional subgroup discovery methods. We provide statistical arguments why this is an improvement. In our experiments, REDS reduces the number of simulations required by 50--75\% on average, depending on the quality measure. It is also useful as a semi-supervised subgroup discovery method and for discovering better scenarios from third-party data, when a simulation model is not available.