Machine Learning in a data-limited regime: Augmenting experiments with synthetic data uncovers order in crumpled sheets

TL;DR

This paper demonstrates that augmenting limited experimental data with synthetic data from a simpler related system enables effective machine learning analysis of complex spatial patterns in crumpled sheets.

Contribution

Introducing a data augmentation strategy using synthetic data from a simpler system to improve machine learning in data-scarce experimental contexts.

Findings

Synthetic data improves pattern prediction accuracy.

Machine learning effectively uncovers order in complex systems.

Augmentation enables analysis with limited experimental data.

Abstract

Machine learning has gained widespread attention as a powerful tool to identify structure in complex, high-dimensional data. However, these techniques are ostensibly inapplicable for experimental systems where data is scarce or expensive to obtain. Here we introduce a strategy to resolve this impasse by augmenting the experimental dataset with synthetically generated data of a much simpler sister system. Specifically, we study spontaneously emerging local order in crease networks of crumpled thin sheets, a paradigmatic example of spatial complexity, and show that machine learning techniques can be effective even in a data-limited regime. This is achieved by augmenting the scarce experimental dataset with inexhaustible amounts of simulated data of rigid flat-folded sheets, which are simple to simulate and share common statistical properties. This significantly improves the predictive power in a test problem of pattern completion and demonstrates the usefulness of machine learning in bench-top experiments where data is good but scarce.