[PFCC Webinar] Professor Yosuke Harashima from Nara Institute of Science and Technology will be presenting ~Material Discovery Using Neural Network Potential~

Professor Yosuke Harashima (Nara Institute of Science and Technology) will give a lecture titled "Material Discovery Using Neural Network Potential."

If you are interested in application examples of neural network potentials and material design that takes structural stability into account, please join us.

Apply here

This webinar is open to not only Matlantis users, but also those who are considering using Matlantis or would like to try it in the future.

However, please note that the content of the lecture does not include an introduction to the basic functions of Matlantis, so you will be able to enjoy the lecture even more if you review the overview of Matlantis (Atomistic simulation tutorial, Chapter 1, Introduction) before attending.

Date and Time	May 20, 2025 16:00-17:00
Speakers	Professor Yosuke Harashima, Nara Institute of Science and Technology Biography 2013: Ph.D. (Science) Osaka University 2013-2019: Postdoctoral Researcher, National Institute for Materials Science, Institute for Solid State Physics, University of Tokyo, National Institute of Advanced Industrial Science and Technology 2019-2020: Specially Appointed Assistant Professor, Nagoya University 2021: Assistant Professor, University of Tsukuba 2022-2024: Assistant Professor, Nara Institute of Science and Technology 2024-Present: Same as above Associate Professor
Venue	Online (Zoom)
Capacity	500 attendees
Participation fee	free
Application Deadline	May 20, 2025 15:00
Application	Apply here
Lecture name	Material Discovery Using Neural Network Potential
Overview of the lecture	In the search for new substances, it goes without saying that various physical properties (catalytic activity, dielectric constant, etc.) are excellent, but the compound must be synthesizable in the first place, and it is important to analyze its structural stability as well as its physical properties. Analysis of structural stability using neural network potentials has a high calculation speed, and it is expected that its application will develop by combining it with other analyses, such as the search for chemical compositions. In this presentation, we propose an analysis that combines a neural network potential with a search for chemical compositions using a generative model and an evaluation of structural stability at finite temperature. In the first half, we introduce the application of a machine learning model that generates chemical compositions that can have specific properties from an infinite number of chemical compositions, and in the second half, we discuss the application of the model to an analysis of the structural stability of random impurity-substituted systems that simultaneously considers lattice vibrations and impurity configurations.

Precautions
We may decline your participation in this webinar at our discretion, such as if we receive an application from a competitor. Thank you for your understanding in advance.