Artificial Intelligence Predicts Body-Centered-Cubic and Face-Centered-Cubic Phases in High-Entropy Alloys

Abstract

For the first time, an artificial intelligence approach, the adaptive neuro fuzzy interface system (ANFIS), has been used for phase predictions in high-entropy alloys. The phase predictions were carried out using two different approaches. The first one was constructed using the composition of elements as the input parameters. The second one was constructed using six parameters, which are found to be crucial in the formation of these alloys. The accuracy of these models was verified using a testing dataset and was found to be 84.21% and 80%, respectively. The discussion demonstrates the application of ANFIS for designing these alloys.