Desenvolvimento de modelo de regressão para previsão do teor de fósforo na etapa de refino primário de aciaria a oxigênio

Abstract

Machine learning is an artificial intelligence area widely applicable. There are many kinds of it, for example linear regression model. Commonly we find machine learning aplications in healthcare systems, public safety, financial services and several others. This papel will evidence an aplication of an algorithm of machine learning that will predict the phosphorus content in the primary refining process, in a steel mill, using linear regression. Data were provided by an steel mills industry with all variables used in this case. Inicially the data base was normalized, then divided in two data packets and finally was implemented a linear regression model. This model presented a partial assertivity, although the adopted metrics have had low values which indicates that is possible an adoption of a predictive models in similar works.