Analysis of the impact of selected factors on the occurrence of corrosion in a specific material using neural networks

Artificial intelligence (AI) technologies are gaining increasing importance in various aspects of life, from daily activities to complex business and scientific processes. The key areas of AI is machine learning and artificial neural networks, which, thanks to their abilities to learn, adapt, and solve complex problems, also open new possibilities in the field of materials science. AI is also used in research on modern materials [1-6], like high-entropy alloys (HEA). The use of artificial neural networks to predict the parameters of these alloys based on their chemical composition can significantly reduce the time and costs associated with laboratory experiments. A predictive model was created using artificial neural networks, which enables the study and analysis of the impact of selected factors on the occurrence of corrosion in a specific material. To achieve this goal, data from laboratory research, scientific literature, empirical knowledge, and guidance from domain experts were used. The predictive model also has the potential for further development and modification, enabling the analysis of the impact of various factors on many mechanical parameters of specific materials. As part of the project, methods of artificial intelligence in the field of artificial neural networks were explored. Existing solutions were analyzed, and a review of scientific literature on related issues was conducted. Then, the analytical part began, in which the dataset used was described and analyzed, forming the basis for further work. The finally developed solution architecture and the created neural network models were used to conduct the final analyses. • Analyzing corrosion susceptibility in the context of chemical composition, temperature, and exposure time on high-entropy alloys (HEA) • Comparison of predictive models in corrosion analysis • Corrosion in HEAs predicted using artificial neural networks and key input features