Predictive Modeling of High-Entropy Alloys and Amorphous\nMetallic Alloys Using Machine Learning

High entropy alloys\nand amorphous metallic alloys represent two\ndistinct classes of advanced alloy materials, each with unique structural\ncharacteristics. Their emergence has garnered considerable interest\nacross the materials science and engineering communities, driven by\ntheir promising properties, including exceptional strength. However,\ntheir extensive compositional diversity poses substantial challenges\nfor systematic exploration, as traditional experimental approaches\nand high-throughput calculations struggle to efficiently navigate\nthis vast space. While the recent development in data-driven materials\ndiscovery could potentially help, such efforts are hindered by the\nscarcity of comprehensive data and the lack of robust predictive tools\nthat can effectively link alloy composition with specific properties.\nTo address these challenges, we have deployed a machine-learning-based\nworkflow for feature selection and statistical analysis to afford\npredictive models that accelerate the data-driven discovery and optimization\nof these advanced materials. Our methodology is validated through\ntwo case studies: (i) a regression analysis of the bulk modulus, and\n(ii) a classification analysis based on glass-forming ability. The\nBayesian-optimized regression model trained for the prediction of\nbulk modulus achieved an <i>R</i><sup>2</sup> of 0.969,\nan mean absolute error (MAE) of 3.958 GPa, and an root mean square\nerror (RMSE) of 5.411 GPa, while our classification model for predicting\nglass-forming ability achieved an F1-score of 0.91, an area-under-the-curve\nof the receiver-operating-characteristic curve of 0.98, and an accuracy\nof 0.91. Furthermore, by leveraging a wide array of chemical data\nfrom diverse literature sources, we have successfully predicted a\nbroad range of properties. This success underscores the efficacy of\nour modeling approach and emphasizes the importance of a comprehensive\nfeature analysis and judicious feature selection strategy over a mere\nreliance on complex modeling techniques.