Molecular simulation-derived features for machine learning predictions of metal glass forming ability

We have developed models of metallic alloy glass forming ability based on newly computationally accessible features obtained from molecular dynamics simulations. In this work we showed that it is possible to increase the predictive value GFA by using input Such require only relatively straightforward and scalable simulations, making them significantly easier less expensive obtain than experimental measurements. generated a database critical cooling rates along with associated candidate are inspired previous research GFA. Out list 9 proposed features, identify two as being most important performance through LASSO model. Enthalpy crystallization icosahedral-like fraction at 100 K promise because they enable significant improvement model flexible ab initio quantum mechanical methods readily applicable almost all systems. This advancement in for machine learning predictions will future more accurately predict new alloys.