Predição de morte de crianças abaixo de 1 ano no estado do paraná

Abstract

This study addresses the importance of utilizing machine learning techniques in the healthcare field, specifically in predicting mortality in children under one year of age. Infant mortality is a significant problem that affects millions of children worldwide and requires an effective approach to reduce these preventable deaths. In this study, machine learning algorithms such as Support Vector Machines (SVM), k-Nearest Neighbors (kNN), and Naive Bayes (NB) were used to develop predictive models. These models were trained based on demographic and healthrelated data collected from a large publicly available dataset. The application of dimensionality reduction techniques, such as the chi-square test and Student’s t-test, allowed for the selection of the most relevant attributes and reduction of dataset complexity. Performance metrics such as accuracy, error rate, sensitivity, specificity, precision, and F1 score were employed to evaluate the models’ performance. Additionally, the area under the receiver operating characteristic curve (AUC-ROC) was used as a performance measure to assess the models’ discrimination capability. The utilization of machine learning techniques in healthcare, such as the prediction of infant mortality, can have a significant impact on resource allocation and the implementation of appropriate interventions. By early identifying risk factors and predicting mortality risk, preventive measures and intervention strategies can be adopted more efficiently. The results of this study can contribute to the understanding of the application of machine learning in healthcare, providing valuable insights for healthcare professionals and aiding in decision-making to improve the health and well-being of children.