Advancements in Surface Behavior Analysis: Engineers Utilize Machine Learning for Automatic Reconstruction
Originally Published 2 years ago — by Phys.org
Engineers at MIT have developed a machine learning-based approach called Automatic Surface Reconstruction that can determine the behavior of material surfaces. This method eliminates the need for intuition and provides more detailed information than conventional methods. By using active learning and a Monte-Carlo algorithm, the system can estimate variations in surface configurations and properties, even under different external conditions. The researchers' tool, AutoSurfRecon, is freely available and can be used to develop new materials for catalysts, hydrogen production, and carbon dioxide removal. The method has already identified new atomic configurations in a material that had been studied for 30 years using traditional methods.