ML potentials

Philipp Marquetand (U of Vienna)

Jul 18. 2022, 09:00 — 10:30

The lectures and tutorials focus on how the potential energy of molecules and materials can be represented by machine learning (ML). The techniques discussed are artificial neural networks and kernel-based methods. Examples of how to represent a compound (representations also called descriptors) in the frame of ML approaches will be given. The first day is centered around neural networks and the tutorial will show how to use the neural network library n2p2 to simulate water. The second day targets kernel-based methods and the tutorial will employ the scikit-learn package to predict energies of a data set called QM7.

Further Information
ESI Boltzmann Lecture Hall
Associated Event:
ESI-DCAFM-TACO-VDSP Summer School on "Machine Learning for Materials Hard and Soft" (Graduate School)
Christoph Dellago (U of Vienna)
Ulrike Diebold (TU Vienna)
Leticia Gonzalez Herrero (U of Vienna)
Jani Kotakoski (U of Vienna)
Christiane Losert-Valiente Kroon (U of Vienna)