- AutorIn
- Felix Conrad Technische Universität Dresden, Dresden, Germany
- Julien Philipp StöckerTechnische Universität Dresden, Dresden, Germany
- Cesare SignoriniTechnische Universität Dresden, Dresden, Germany
- Isabela de Paula Salgado
- Hajo Wiemer
- Michael Kaliske
- Steffen Ihlenfeldt
- Titel
- Exploring design space
- Untertitel
- Machine learning for multi-objective materials design optimization with enhanced evaluation strategies
- Zitierfähige Url:
- https://nbn-resolving.org/urn:nbn:de:bsz:14-qucosa2-1000939
- Quellenangabe
- Computational materials science
Erscheinungsjahr: 2025
Jahrgang: 246
E-ISSN: 1879-0801
Artikelnummer: 113432 - Erstveröffentlichung
- 2024
- Abstract (EN)
- Discovering optimal material designs in the design space can be significantly accelerated by leveraging machine learning (ML) models for screening candidates. However, the quality of these designs depends on the prediction accuracy of the ML models and the efficiency of the optimization algorithms used. This study comprehensively compares different ML modeling strategies, optimization algorithms and evaluation strategies. Thereby, automated ML, tree-based ML models and neural networks were compared. Various optimization algorithms were analyzed, including random search, evolutionary and swarm-based methods. In addition, different strategies for evaluating the predictive performance of the ML models were investigated, which is particularly important as these models are expected to predict design parameters that deviate significantly from the known designs in the training data throughout the optimization. Our results highlight the capability of the proposed workflow to discover material designs that significantly outperform those within the training database and approach theoretical optima. Overall, this research contributes to advancing the field of material design optimization by providing a versatile and practical workflow that introduces automated ML into material design optimization and new model error assessment strategies tailored explicitly to optimization tasks.
- Andere Ausgabe
- Link zum Artikel, der zuerst in der Zeitschrift „Computational materials science” im Verlag Elsevier erschienen ist.
DOI: 10.1016/j.commatsci.2024.113432 - Verweis
- Ergänzendes Material ist unter folgendem Link zu finden.
Link: https://www.sciencedirect.com/science/article/pii/S0927025624006530?via%3Dihub#appA - Freie Schlagwörter (EN)
- Machine learning in materials design, Automated machine learning, Multi-objective optimization, Splitting methods for performance evaluation
- Klassifikation (DDC)
- 530
- Verlag
- Elsevier, Amsterdam
- Förder- / Projektangaben
- Deutsche Forschungsgemeinschaft (DFG)
Graduiertenkollegs
Impaktsicherheit von Baukonstruktionen durch mineralisch gebundene Komposite
(GRK 2250)
ID: 287321140 - Deutsche Forschungsgemeinschaft (DFG)
TRR 339: Digitaler Zwilling Straße – Physikalisch-informatorische Abbildung des Systems „Straße der Zukunft“
Physikbasierte neuronale Netze für Langzeitprognosen von Fahrbahnen und Unschärfemodellierung (B04)
ID: 453596084 - Bundesministrium für Bildung und Forschung (BMBF)
Ganzheitliche KI-basierte Optimierung von Kunststoffverpackungen mit Rezyklatanteil
(KIOptiPack)
ID: 033KI129 - Version / Begutachtungsstatus
- publizierte Version / Verlagsversion
- URN Qucosa
- urn:nbn:de:bsz:14-qucosa2-1000939
- Veröffentlichungsdatum Qucosa
- 26.10.2026
- Dokumenttyp
- Artikel
- Sprache des Dokumentes
- Englisch
- Lizenz / Rechtehinweis
CC BY 4.0