Quantum support vector classifier for phase diagram prediction in quinary systems

Abstract

The integration of machine learning (ML) in materials science has accelerated the discovery and optimization of novel materials. However, classical ML approaches often face limitations in handling the increasing complexity and scale of modern datasets. Quantum machine learning (QML), leveraging quantum computing principles, offers a promising avenue to address these challenges. This study explores the application of the quantum support vector classifier (QSVC) to predict phase diagrams in the Al–Cu–Mg–Si–Zn quinary system. The prediction of the phase diagram of quinary systems contributes to the achievement of issues related to sustainable development goals by facilitating the creation of sophisticated materials that possess exceptional characteristics. This, in turn, promotes innovation and sustainable practices within the industrial sector. We used a comprehensive dataset from high-throughput CALPHAD calculations, and employed QSVC with advanced quantum feature transformations and kernel methods. Our results demonstrate significant improvements in predictive accuracy and efficiency compared to classical SVC, highlighting the potential of QML to advance material design.