001     917552
005     20240712112854.0
024 7 _ |a 10.48550/ARXIV.2207.13779
|2 doi
024 7 _ |a 2128/33643
|2 Handle
037 _ _ |a FZJ-2023-00754
100 1 _ |a Schweidtmann, Artur M.
|0 P:(DE-HGF)0
|b 0
245 _ _ |a Physical Pooling Functions in Graph Neural Networks for Molecular Property Prediction
260 _ _ |c 2022
|b arXiv
336 7 _ |a Preprint
|b preprint
|m preprint
|0 PUB:(DE-HGF)25
|s 1673946218_28073
|2 PUB:(DE-HGF)
336 7 _ |a WORKING_PAPER
|2 ORCID
336 7 _ |a Electronic Article
|0 28
|2 EndNote
336 7 _ |a preprint
|2 DRIVER
336 7 _ |a ARTICLE
|2 BibTeX
336 7 _ |a Output Types/Working Paper
|2 DataCite
520 _ _ |a Graph neural networks (GNNs) are emerging in chemical engineering for the end-to-end learning of physicochemical properties based on molecular graphs. A key element of GNNs is the pooling function which combines atom feature vectors into molecular fingerprints. Most previous works use a standard pooling function to predict a variety of properties. However, unsuitable pooling functions can lead to unphysical GNNs that poorly generalize. We compare and select meaningful GNN pooling methods based on physical knowledge about the learned properties. The impact of physical pooling functions is demonstrated with molecular properties calculated from quantum mechanical computations. We also compare our results to the recent set2set pooling approach. We recommend using sum pooling for the prediction of properties that depend on molecular size and compare pooling functions for properties that are molecular size-independent. Overall, we show that the use of physical pooling functions significantly enhances generalization.
536 _ _ |a 1121 - Digitalization and Systems Technology for Flexibility Solutions (POF4-112)
|0 G:(DE-HGF)POF4-1121
|c POF4-112
|f POF IV
|x 0
588 _ _ |a Dataset connected to DataCite
650 _ 7 |a Machine Learning (cs.LG)
|2 Other
650 _ 7 |a FOS: Computer and information sciences
|2 Other
700 1 _ |a Rittig, Jan G.
|0 P:(DE-HGF)0
|b 1
700 1 _ |a Weber, Jana M.
|0 P:(DE-HGF)0
|b 2
700 1 _ |a Grohe, Martin
|0 P:(DE-HGF)0
|b 3
700 1 _ |a Dahmen, Manuel
|0 P:(DE-Juel1)172097
|b 4
|u fzj
700 1 _ |a Leonhard, Kai
|0 P:(DE-HGF)0
|b 5
700 1 _ |a Mitsos, Alexander
|0 P:(DE-Juel1)172025
|b 6
|e Corresponding author
|u fzj
773 _ _ |a 10.48550/ARXIV.2207.13779
856 4 _ |u https://juser.fz-juelich.de/record/917552/files/2207.13779.pdf
|y OpenAccess
909 C O |o oai:juser.fz-juelich.de:917552
|p openaire
|p open_access
|p VDB
|p driver
|p dnbdelivery
910 1 _ |a RWTH Aachen
|0 I:(DE-588b)36225-6
|k RWTH
|b 1
|6 P:(DE-HGF)0
910 1 _ |a Forschungszentrum Jülich
|0 I:(DE-588b)5008462-8
|k FZJ
|b 4
|6 P:(DE-Juel1)172097
910 1 _ |a RWTH Aachen
|0 I:(DE-588b)36225-6
|k RWTH
|b 5
|6 P:(DE-HGF)0
910 1 _ |a Forschungszentrum Jülich
|0 I:(DE-588b)5008462-8
|k FZJ
|b 6
|6 P:(DE-Juel1)172025
910 1 _ |a RWTH Aachen
|0 I:(DE-588b)36225-6
|k RWTH
|b 6
|6 P:(DE-Juel1)172025
913 1 _ |a DE-HGF
|b Forschungsbereich Energie
|l Energiesystemdesign (ESD)
|1 G:(DE-HGF)POF4-110
|0 G:(DE-HGF)POF4-112
|3 G:(DE-HGF)POF4
|2 G:(DE-HGF)POF4-100
|4 G:(DE-HGF)POF
|v Digitalisierung und Systemtechnik
|9 G:(DE-HGF)POF4-1121
|x 0
914 1 _ |y 2022
915 _ _ |a OpenAccess
|0 StatID:(DE-HGF)0510
|2 StatID
920 _ _ |l yes
920 1 _ |0 I:(DE-Juel1)IEK-10-20170217
|k IEK-10
|l Modellierung von Energiesystemen
|x 0
980 1 _ |a FullTexts
980 _ _ |a preprint
980 _ _ |a VDB
980 _ _ |a UNRESTRICTED
980 _ _ |a I:(DE-Juel1)IEK-10-20170217
981 _ _ |a I:(DE-Juel1)ICE-1-20170217


LibraryCollectionCLSMajorCLSMinorLanguageAuthor
Marc 21