PhD candidate
Research group | Hylleraas Centre for Quantum Molecular Sciences RT5
Main supervisor | David Balcells
Co-supervisors | R. De Bin & T.B. Pedersen
Affiliation | Department of Chemistry, UiO
Contact | hanneskn@uio.no
Short bio
My academic background is computational physics with a focus on computational statistics and quantum chemistry. In my master thesis I worked on genetic algorithms for optimizing molecular properties in a given chemical space.
Research interests and hobbies
My main research interests lie in the development and employment of predictive as well as generative machine learning models for chemical applications based on quantum chemistry data.
CompSci project
Catalyst discovery by combining computational chemistry with machine learning
The starting point of this project is the tmQM data set which consists of approximately 86k data points of transition metal complexes, each featuring electronic structure data and quantum property data. Using this data set machine learning approaches shall be applied in order to build predictive as well as generative models for transition metal complexes. One of the main aspects therein is the generation of appropriate representations of molecules in terms of graphs that incorporate the relevant physics in an efficient way. These graph representations in turn call for special neural network architectures (message passing neural networks) that can handle such structurally diverse data structures. Corresponding preliminary work for organic molecules exists and one open research question is how the performance for transition metal complexes compares to these results.
Publications
CompSci publications
- Hannes Kneiding, Ruslan Lukin, Lucas Lang, Simen Reine, Thomas Bondo Pedersen, Riccardo De Bin and David Balcells (2023) “Deep learning metal complex properties with natural quantum graphs” Digital Discovery 2 (3) 618–633
https://doi.org/10.1039/D2DD00129B | Full text in Research Archive - Hannes Kneiding, Ainara Nova and David Balcells (2024) “Directional multiobjective optimization of metal complexes at the billion-system scale” Nature Computational Science 4, 263-273
https://doi.org/10.1038/s43588-024-00616-5 | Full text in Research Archive
Previous publications
None yet.
