Publications of Johannes Margraf
All genres
Talk (30)
2023
Talk
Margraf, Johannes: Science Driven Chemical Machine Learning.
(12th SolTech Conference 2023, Würzburg, Germany, Oct 2023).
Talk
Margraf, Johannes: A Personal Perspective on ML Interatomic Potentials.
(Crash TEsting machine learning force fields: Applicability, best practices, limitations (TEA 2023), Luxembourg, Luxembourg, Oct 2023).
Talk
Margraf, Johannes: Science Driven Chemical Machine Learning.
(Thomas Young Center-FHI Workshop, London, UK, Feb 2023).
Talk
Margraf, Johannes: Science Driven Chemical Machine Learning.
(Joint Seminar of Theory and Computational Chemistry, Erlangen, Germany, Nov 2023).
Talk
Margraf, Johannes: Discovering Molecules and Materials With Machine Learning.
(Seminar, Research Center for Modeling and Simulation, MODUS, Bayreuth, Germany, Dec 2023).
2022
Talk
Margraf, Johannes: Heterogeneous Catalysis in Grammar School.
(FHI-Workshop on Current Research Topics at the FHI, Potsdam, Germany, May 2022).
Talk
Margraf, Johannes: Data-Efficient Chemical Machine Learning.
(KAIST Theory Seminar, Seoul, South Korea, Online Event, Jan 2022).
Talk
Margraf, Johannes: Data-Efficient Chemical Machine Learning.
(Institutskolloquium, Institute of Chemistry, University of Potsdam, Online Event, Jan 2022).
Talk
Margraf, Johannes: Describing Complex Polar Materials With Physics-Enhanced Machine Learning.
(ACS Spring Meeting 2022, Symposium, Complexity in Computational Catalysis: Balancing Model and Method Accuracy: Machine Learning and Kinetic Modeling, Online Event, Mar 2022).
Talk
Margraf, Johannes: Predicting Molecular Properties through Machine Learned Energy Functionals.
(ML4M 2022, Young Researcher’s Workshop on Machine Learning for Materials 2022, Trieste, italy, May 2022).
Talk
Margraf, Johannes: Predicting Molecular Properties through Machine Learned Energy Functionals.
(Seminar, VirtMat, Karlsruhe Institute of Technology (KIT), Online Event, Jun 2022).
Talk
Margraf, Johannes: ∆-Learning with DFTB: What makes a good baseline?
(Workshop, Multi-Scale Quantum Mechanical Analysis of Condensed Phase Systems: Methods and Applications, Telluride Science Research Center, Telluride, CO, USA, Jul 2022).
Talk
Margraf, Johannes: Science Driven Chemical Machine Learning.
(Seminar, Debye Institute for Nanomaterials Science, Utrecht University, Utrecht, The Netherlands, Nov 2022).
2021
Talk
Margraf, Johannes: Chemical ML Beyond Established Benchmark Datasets.
(Workshop, ELLIS Machine Learning (ML) for Molecule Discovery, Online Event, Dec 2021).
Talk
Margraf, Johannes: Predicting Molecular Properties Through Machine Learned Energy Functionals.
(Discussion Meeting, GdR REST Machine Learning (ML), Online Event, Dec 2021).
Talk
Margraf, Johannes: Integrating Machine Learning and Electronic Structure Theory.
(FHI-Workshop on Current Research Topics at the FHI, Online Event, Jun 2021).
2020
Talk
Margraf, Johannes: Molecular Machine Learning: From Chemical Space to Reaction Space.
(FHI-Workshop on Current Research Topics at the FHI, Online Event, May 2020).
Thesis - PhD (1)
2025
Thesis - PhD
Cui, Mengnan: Density Functional Tight Binding Theory Informed Multi-fidelity Machine Learning.
Working Paper (2)
2025
Working Paper
Jakob, Konstantin, Karsten Reuter and Johannes Margraf: Universally Accurate or Specifically Inadequate? Stress-testing General Purpose Machine Learning Interatomic Potentials.
2021
Working Paper
, Johannes Margraf, Karsten Reuter and : Interplay between shape and composition in bimetallic nanoparticles revealed by an efficient optimal-exchange optimization algorithm.