This repository contains the dataset and code for my paper Análise da informatividade de microgrupos em Mapas Auto-Organizáveis para identificação de variáveis importantes no diagnóstico de COVID-19.

This paper describes a study based on the classification results of a machine learning algorithm in relation to hemograms of patients with suspected COVID-19. Based on the hypothesis that the most important variables in the blood count are those that have less informativeness in Self-Organizing Maps, we sought to identify, through the SOMLI-KNN algorithm, the informativeness metric and perform microgroup analysis of the Self-Organizing Map so that this hypothesis can be validated. As a result, the new approach showed rapid execution and high accuracy in classifying the diagnosis of patients with suspected COVID-19 compared to other classifiers in the literature.

Análise da informatividade de microgrupos em Mapas Auto-Organizáveis para identificação de variáveis importantes no diagnóstico de COVID-19

The dataset folder in this repository contains the dataset used in this paper. COVID-19 diagnostic tests data obtained from Instituto Fleury were used as the dataset.

Vinícius Gomes Pajaro Grande, Leandro Augusto da Silva