||This symposium will provide a forum to identify current achievements and existing challenges in the modeling of ceramic and glassy materials by simulation and machine learning. The symposium will span over various types of materials (ordered or disordered), lengthscales, and properties. Computational techniques of interest include: classical and ab initio molecular dynamics simulations, mesoscale simulations, continuum modeling, data mining, machine learning, natural language processing, optimization, etc. Contributions that beneficially combine physics-based simulations and machine learning (wherein one technique informs, advances, or replaces the other) are of special interest, but studies solely focusing on simulation or machine learning are also welcomed.
Topics include, but are not limited to:
• Informatics and machine learning to predict materials properties,
• Physics-informed machine learning,
• Data mining and automated data extraction from the literature,
• Interatomic forcefield development,
• Machine learning approaches to decode structure-property relationships,
• High-throughput simulations to generate big data,
• First-principle and classical modeling for structure and property prediction,
• Upscaling techniques and mesoscale modeling,
• Continuum modeling of glasses and ceramic materials,
• Meta-optimization and inverse design, and,
• Machine learning for image/microstructure analysis.