| Beschreibung |
This project-based course focuses on the integration of AI in material science, by utilizing machine learning and deep learning to model and design materials with targeted mechanical properties. Students gain hands-on knowledge towards: - Building predictive models
- Comparing feature encodings
- Performing sensitivity analysis
- Evaluating performance under limited data, uncertainty and out-of-distribution conditions
This course progresses from classical machine learning models to ensemble methods and deep neural networks, with an emphasis on representation learning, uncertainty-aware modeling and robustness. In the final phase, students will apply their trained models to inverse design problems to search for material compositions that meet target performance constraints. The final outcome includes a reproducible code repository, a technical report, and a final presentation summarizing methods, results, and design insights. |
| Literatur |
1] I.-C. Yeh, ”Modeling of strength of high-performance concrete using artificial neural networks,” Cement and Concrete Research, vol. 28, pp. 1797–1808, 1998. [2] U. J. Malik, C. K. Lee, D. Mohotti, and H. Mo, ”Global Dataset of UHPC Mix Designs with Supplementary Cementitious Materials, Nano Additives, and Sustainable Fillers,” Mendeley Data, V2, 23 Sep. 2025. |
