11:45 AM - 12:15 PM
[8a-Z28-5] Exploring diverse stable phases by pseudo-free-energy: Information analysis of domain structure
Keywords:Machine learning, Magnetic domain, Diverse stable phases
We aim to establish a relational connection between micro-structure and macro-function by integrating materials and information sciences. We here propose a new analysis method using a "pseudo-free energy landscape". Microstructure are quantified by Persistent homology for feature extraction. The energy landscape is modeled using explainable machine learning methods. This is expected to lead to deep analysis of the interpretation of functions. The presentation will include several case studies.