Predicting the Inhibition Efficiencies of Magnesium Dissolution Modulators using Sparse Machine Learning Models
Abstract
The degradation behaviour of magnesium and its alloys can be tuned by small organic molecules. However, an automatic identification of effective organic additives within the vast chemical space of potential compounds needs sophisticated tools. Herein, we propose two systematic approaches of sparse feature selection for identifying molecular descriptors that are most relevant for the corrosion inhibition efficiency of chemical compounds. One is based on the classical statistical tool of analysis of variance, the other one based on random forests. We demonstrate how both can—when combined with deep neural networks—help to predict the corrosion inhibition efficiencies of chemical compounds for the magnesium alloy ZE41. In particular, we demonstrate that this framework outperforms predictions relying on a random selection of molecular descriptors. Finally, we point out how autoencoders could be used in the future to enable even more accurate automated predictions of corrosion inhibition efficiencies.
Elisabeth Schiessler
Doktorandin
Genetische Algorithmen, Maschinelles Lernen im Bereich Materialwissenschaften, Autoencoder
Tim Würger
Doktorand
Computergestützte Materialwissenschaften, Atomistische Simulation & Modellierung, Maschinelles Lernen und datengetriebene Ansätze für Materialmodellierung & -design
Robert Meißner
Professor
Molekulardynamische Simulationen von Grenzflächen, Berechnung freier Energien biomolekularer und elektrochemischer Systeme, atomistische Betrachtung der Magnesiumkorrosion, Entwicklung datengetriebener Modelle zur Indentifikation von (umweltfreundlichen) Degradationsmodulatoren
Christian Cyron
Professor
Computational Engineering, Maschinelles Lernen in Maschinenbau, Materialforschung und Medizin
Mikhail Zheludkevich
Professor
Elektrochemie, Multifunktionale Oberflächen & Aktiver Schutz von Leichtmetallen, Multi-Material-Systeme
Christian Feiler
Postdoc & MLE-Koordinator
Atomistische Simulation & Modellierung, Entwicklung von Quantitativen Struktur-Eigenschafts-Beziehungsmodellen