Author(s)

Haiyang Dong

Corresponding Author

Haiyang Dong

Abstract

The neural network ontology model is a useful tool for explaining the mechanical behavior of complicated materials because it combines the ontological relationship of materials science with the potent nonlinear fitting ability of neural networks. The use of neural networks in materials science has gradually expanded in recent years due to the quick development of computing power and data processing technology. This is particularly true for real-time prediction, large-scale data processing, and complex materials modeling, all showing great promise. Nevertheless, there are still a lot of issues that neural network ontology models must overcome in real-world settings, including large data requirements, poor model interpretability, insufficient generalization capability, and high processing costs. The purpose of this study is to serve as a reference and guide for future research in this paper by reviewing the fundamental concepts of the neural network ontology model, highlighting its primary application areas and obstacles, and suggesting research options.

Keywords

Neural network; Ontology model; Material behavior; Data driven; Physics-informed neural networks