Sanggyu Chong^1*, Filippo Bigi^1*, Federico Grasselli^1*, Philip Loche^1*, Matthias Kellner^1*, Michele Ceriotti^1*

1 Laboratory of Computational Science and Modeling (COSMO), IMX, École Polytechnique Fédérale de Lausanne (EPFL), 1015 Lausanne, Switzerland

* Corresponding authors emails: sanggyu.chong@epfl.ch, filippo.bigi@epfl.ch, federico.grasselli@epfl.ch, philip.loche@epfl.ch, matthias.kellner@epfl.ch, michele.ceriotti@epfl.ch

DOI10.24435/materialscloud:6x-gs [version v1]

Publication date: Aug 28, 2024

How to cite this record

Sanggyu Chong, Filippo Bigi, Federico Grasselli, Philip Loche, Matthias Kellner, Michele Ceriotti, Prediction rigidities for data-driven chemistry, Materials Cloud Archive 2024.130 (2024), https://doi.org/10.24435/materialscloud:6x-gs

Description

The widespread application of machine learning (ML) to the chemical sciences is making it very important to understand how the ML models learn to correlate chemical structures with their properties, and what can be done to improve the training efficiency whilst guaranteeing interpretability and transferability. In this work, we demonstrate the wide utility of prediction rigidities, a family of metrics derived from the loss function, in understanding the robustness of ML model predictions. We show that the prediction rigidities allow the assessment of the model not only at the global level, but also on the local or the component-wise level at which the intermediate (e.g. atomic, body-ordered, or range-separated) predictions are made. We leverage these metrics to understand the learning behavior of different ML models, and to guide efficient dataset construction for model training. We finally implement the formalism for a ML model targeting a coarse-grained system to demonstrate the applicability of the prediction rigidities to an even broader class of atomistic modeling problems. This record contains all the data used for the analyses conducted in the associated work published in Faraday Discussions.

Materials Cloud sections using this data

Files

File name	Size	Description
PR_FD_materials_cloud.tar.gz MD5md5:4e5ab9528477908af3c68637ef0d1154	4.5 GiB	compressed file containing the reference structures and their associated data, as well as trained NN models
README.txt MD5md5:5952c12112a7efa01dc6b6bf2bde36c6	2.0 KiB	README file with details about the data, how it has been organized, and where additional information about how to access the trained models can be found

License

Files and data are licensed under the terms of the following license: Creative Commons Attribution 4.0 International.
Metadata, except for email addresses, are licensed under the Creative Commons Attribution Share-Alike 4.0 International license.

External references

Keywords

machine learning prediction rigidity uncertainty quantification ML model robustness