Título: Predição de Corrosão Atmosférica em Materiais Metálicos Utilizando Aprendizado de Máquina
Autores: Vinicius Michelon Geremias, Thiago Brandenburg, Gustavo A. Fischer, Fabiano Miranda, Jose Francisco Silva Filho, Rafael Stubs Parpinelli
Resumo: The study of the corrosion impact on metallic materials is of great importance to industries and the metallurgical market. Over the years, several formulas, known as Dose-Response Functions, have been developed with the aim of predicting corrosion based on environmental factors. Using data from the MICAT atmospheric corrosion study, this article proposes data-driven prediction models for four materials, named Carbon Steel, Aluminum, Copper, and Zinc. The models indicate that the Random Forest algorithm can be used to predict atmospheric corrosion of metallic materials with competitive results when compared to standard prediction functions
Palavras-chave: Random Forest, Atmospheric Corrosion, MICAT
Páginas: 8
Código DOI: 10.21528/CBIC2023-101
Artigo em pdf: CBIC_2023_paper101.pdf
Arquivo BibTeX: CBIC_2023_101.bib