Machine Learning Enables Scientists Find New Materials With Desirable Properties

Machine learning has been harnessed to find new materials with specific properties. Scientists from the Los Alamos National Laboratory have used an informatics-based adaptive strategy in a novel approach to enable scientists to find materials that are more cost-effective and less time consuming.

"What we've done is to show that starting with a relatively small data set of well-controlled experiments, it is possible to iteratively guide subsequent experiments toward finding the material with the desired target," said Turab Lookman, a physicist and materials scientist at Los Alamos National Laboratory and senior author of the study.

"Finding new materials has traditionally been guided by intuition and trial and error," he continued. "But with increasing chemical complexity, the combination possibilities become too large for trial-and-error approaches to be practical."

The tedious framework  can guide experiments to identify a shape-memory alloy with low dissipation. This property is significant to enhance fatigue life in engineering applications.

"The goal is to cut in half the time and cost of bringing materials to market," Lookman said. "What we have demonstrated is a data-driven framework built on the foundations of machine learning and design that can lead to discovering new materials with targeted properties much faster than before."

The machine-learning algorithm is powered by Los Alamos National Laboratory's supercomputers. The entire "trial-and-error process" has been digitized through the supercomputers.

"Using the interplay of structural, chemical and microstructural degrees of freedom allowed the team to introduce a great degree of flexibility in comparison to standard procedures that create and screen databases using thousands of quantum mechanical calculations and do not incorporate uncertainties."

This novel approach can be used for any kind of material from polymers and ceramics to nanomaterials.

The findings were published on April 15, 2016, issue of Nature Communications.