Fig. 1: The schematics for the development of our modeling-guided hybrid ML algorithm.
From: Rational design of chemically complex metallic glasses by hybrid modeling guided machine learning
a The comparison of different types of MGs in terms of their proportion in the hitherto reported MGs with measured GFAs. The inset highlights the elements used in the prior development of MGs with counts indicating the total number of times of an individual element being found in the reported MG compositions. b The breakdown of the data descriptors we developed. c The illustration for the training/validation of our classification ML model based on adaptive boosting (AB), support vector machine (SVM), and k-nearest neighbor (KNN). d The illustrated four types of regression ML models. e The illustration of the predicted GFA diagrams for ternary alloys. f The development of MGs through ML predictions. The scale bar indicates a length of 5 mm.
