Public Thesis defense - IMCN

Machine Learning topological characteristics from multiple electronic materials databases by Yuqing HE

Pour l’obtention du grade académique de Doctorat en sciences de l’ingénieur et technologie

Significant advances have been made in predicting new topological materials using high-throughput ab–initio and symmetry-based indicator calculations. To date, thousands of materials have been identified to be topologically nontrivial by scanning the existing databases. However, this approach is severely limited to non-magnetic systems with well-defined symmetries and is not yet able to map to the basic factors that can be applied in the experimental field, leaving a much larger materials space unexplored. We apply the machine learning method and explore the topological materials data to overcome the above shortcomings. A large dataset including 35,608 entries is obtained by merging two symmetry-indicator-based databases: Materiae and Topological Materials Database. We trained a model that reached an accuracy of 85.2% to explore new materials and studied the essential factors that influence the topology of materials by interpreting important features used in our model.

Jury members : 

• Prof. Gian-Marco Rignanese (UCLouvain), supervisor
• Dr. Matteo Giantomassi (UCLouvain), supervisor
• Prof. Xavier Urbain (UCLouvain), chairperson
• Prof. Xavier Gonze (UCLouvain), secretary
• Prof. Jean-Christophe Charlier (UCLouvain)
• Prof. Hongming Weng (Chinese Academy of Sciences, China)
• Prof. Xiaotong Liu (Bejing Information Science and Technology University, China)

Pay attention :

The public defense of Yuqing He scheduled for Thursday 07 December at 04:15 p.m will also take place in the form of a video conference

Télécharger l'annonce