Domain Knowledge-Guided Machine Learning
This presentation first briefly introduces the concept of materials/mechanics informatics. Materials/mechanics informatics is growing extremely fast by integrating machine learning with materials/mechanics science and engineering, where techniques, tools, and theories drawn from the emerging fields such as data science, internet, computer science and engineering, and digital technologies, are applied to the materials/mechanics science and engineering to accelerate materials/mechanics, products and manufacturing innovations.
Then, this presentation reports a domain knowledge-guided machine learning strategy and demonstrate it by studying the oxidation behaviours of ferritic-martensitic steels in supercritical water. This strategy leads to the development of a formula with high generalization and accurate prediction power, which is most desirable to science, technology, and engineering.