Evolução das doenças crônicas: um estudo de ciência de dados aplicado à saúde pública

Abstract

This study proposes the application of data science techniques to analyze chronic kidney disease mortality over time, considering the influence of factors such as Diabetes Mellitus and cardiovascular diseases. Based on a historical dataset containing causes of death by country and year, a process of data cleaning, exploratory analysis, and implementation of two statistical models was carried out: multiple linear regression and second-degree polynomial regression. Both models were evaluated using performance metrics such as R², MAE, MSE, and RMSE. The results showed that the polynomial regression model performed best, with an R² of 0.9999 and a mean absolute error of only 1,645 deaths, while the linear model yielded an R² of 0.9975 and a MAE of 11,287 deaths. The analysis of the coefficients and model structure indicated a strong correlation between Diabetes Mellitus and the progression of chronic kidney disease, also capturing the curvatures present in the temporal data. It is concluded that the application of data science techniques, through statistical modeling and predictive evaluation, is a valuable tool for investigating complex patterns in public health data, providing quantitative evidence to support the formulation of preventive policies and chronic disease monitoring strategies. Keywords: data science, linear regression, polynomial regression, chronic kidney disease, Diabetes Mellitus, cardiovascular diseases.