WW Kolloquium: Prof. Dr. Luca M. Ghiringhelli – From FAIR data to materials modeling: artificial-intelligence methods and data-base infrastructure
Date: 21. November 2023Time: 16:00 – 17:30Location: H14 / hybrid
Prof. Dr. Luca M. Ghiringhelli
Group of Data-Based Materials Modeling, Chair for Materials Simulation
Department of Materials Science and Engineering
Friedrich-Alexander University , Erlangen
Inaugural Lecture: From FAIR data to materials modeling: artificial-intelligence methods and data-base infrastructure
To accelerate the identification and design of optimal materials for a desired property or process, strategies for a well-guided exploration of the materials space are highly needed. A desirable strategy would be to start from experimental or theoretical data, and by means of artificial-intelligence (AI), to identify yet unseen patterns in the data, and consequentially predictive data-driven models. This leads to the identification of materials' (properties) maps. I will present novel methods based on symbolic inference, and their recent updates, for the AI-aided identification of descriptors and materials maps, tailored to work (also) with small-data, and applied to important materials-science challenges such as the prediction of mechanical properties of perovskite materials, of catalytic properties of experimentally characterized materials, and more.
In order to train AI models, well-annotated data are a necessity and the availability of re-usable data generated by the community would be highly convenient. In other words, the data for AI modeling should be FAIR (Findable, Accessible, Interoperable, Reusable), a popular acronym which could also be recast into Fair & AI-Ready. I will introduce the FAIRmat consortium and the NOMAD infrastructure, for the FAIR storage and stewardship of materials-science (meta)data. I will focus on the AI-toolkit, an online platform for publishing and sharing curated Jupyter notebooks, designed for the tutorial introduction of text-book and novel AI tools and for providing an interactive access to AI workflows as published in peer-reviewed journals. In this way, scientists can fully benefit of the community's advancements and scientific reproducibility can meet its full potential.
H14 / hybrid