Quantum Machine Learning in Materials Prediction: A Case Study on ABO₃ Perovskite Structures.

Area of Science:

• Materials Science
• Quantum Computing
• Machine Learning

Background:

• Materials discovery and screening are crucial for scientific advancement.
• Quantum machine learning (QML) presents a novel computational paradigm for complex scientific problems.
• Identifying specific material structures, like perovskites, is essential for various applications.

Purpose of the Study:

• To introduce and evaluate a hybrid classical-quantum machine learning model for classifying ABO3 compounds.
• To assess the efficacy of a variational quantum classifier in identifying simple perovskite structures.
• To demonstrate the potential of QML in materials science classification tasks with limited datasets.

Main Methods:

• Development of a hybrid classical-quantum machine learning model.
• Utilized a variational quantum classifier trained on a dataset of 397 ABO3 compounds.
• Employed feature correlation analysis to optimize the QML system.

Main Results:

• The QML system achieved an 88% accuracy on training data.
• An accuracy of 87% was obtained on unseen test data.
• The model successfully identified simple perovskite structures within the ABO3 dataset.

Conclusions:

• QML, particularly hybrid approaches, shows significant potential for materials science classification.
• The study highlights the effectiveness of QML even with limited training data.
• Quantum computation offers enhanced capabilities for materials investigation and discovery.