This paper introduces a generative deep-learning framework (GDLF) for designing compositionally complex bulk metallic glasses (BMGs), such as high entropy BMGs. The framework uses a Generative Adversarial Network (GAN) for data generation and a supervised Boosted Trees algorithm for evaluation. The study systematically investigates the impact of various data descriptors and demonstrates the framework's ability to produce composition-property mappings, enabling inverse design of BMGs. The framework successfully generated novel BMG compositions, which were experimentally validated.
Publisher
npj Computational Materials
Published On
Jan 23, 2023
Authors
Ziqing Zhou, Yinghui Shang, Xiaodi Liu, Yong Yang
Tags
Generative Adversarial Network
bulk metallic glasses
high entropy BMGs
composition-property mapping
inverse design
data descriptors
deep learning
Related Publications
Explore these studies to deepen your understanding of the subject.
