Informatics and Data Analytics in Ceramics and Glass

Co-lead organizers: Jessica Rimsza (jrimsza@sandia.gov, Sandia National Laboratories, U.S.), Arrigo Calzolari (arrigo.calzolari@nano.cnr.it, CNR Nano Center, Italy), and Stefano Curtarolo (stefano.curtarolo@duke.edu, Duke University, U.S.)

Abstract: This symposium will focus on current achievements and challenges in the modeling of ceramics and glasses through simulation, informatics, and machine learning. The symposium will span various material types, length scales, time scales, and properties. Investigations performed using various computational techniques are of interest in this symposium, including classical and ab initio molecular dynamics simulations, mesoscale simulations, continuum modeling, data mining, machine learning, natural language processing, optimization, and others. Contributions that combine physics-based simulations and machine learning or informatics are of special interest, but studies focusing on simulation or machine learning are also encouraged to submit.

Proposed sessions:

• Informatics and machine learning for prediction of materials properties
• Machine learning approaches to identify structure–property relationships
• Physics-informed machine learning for ceramics and glasses
• Development of interatomic forcefields via machine learning
• High-throughput simulations to generate big data for informatics
• First-principle and classical modeling for structure and property prediction
• Mesoscale and continuum modeling of glasses and ceramic materials
• Machine learning for image/microstructure analysis

Co-organizers:

Bikramajit Basu, IISC Bangalore, India, bikram@iisc.ac.in

Adama Tandia, Corning, U.S. tandiaa@corning.com