Modelo logístico y redes neuronales para pronóstico de anemia en menores de 5 años en el hospital pediátrico alfonso villagómez román periodo 2020-2021

Abstract

The main objective of the present study was to determine the best model for predicting anemia in children under 5 years of age treated at the Alfonso Villagómez-Román Pediatric Hospital during the period 2020-2021. First, a comprehensive review of the literature was performed to identify the possible factors associated with childhood anemia, which allowed a relevant theoretical framework for analysis to be established. Subsequently, relevant information on these factors was collected by accessing the clinical records of children under 5 years of age with anemia who were treated at the pediatric hospital. With the data set collected, we proceeded to model the presence of anemia in children under 5 years of age using both the Logistic Model and the Neural Network Model. Both models were applied and meticulously evaluated through goodness of fit such as the confusion matrix and ROC curve metrics to obtain accurate forecasts. The results obtained were compared to determine which of the two models offered better predictive ability. In this comparative analysis, it was found that Neural Networks outperformed the Logistic Model, proving to be the most effective prediction model for predicting anemia in children under 5 years of age. In conclusion, the study demonstrates that the use of Neural Networks for the prognosis of anemia in children under 5 years of age is more effective than the Logistic Model. These findings have significant implications for pediatric health care, as accurate prognosis of anemia in early stages can contribute to timely medical intervention and improve child health care