Modelos predictivos en la clasificación de donantes a plaquetoféresis sanguínea en el Hospital Nacional Edgardo Rebagliati Martins, Lima-Perú. 2022

The objective of the present investigation was to determine the predictive model that allows classifying donors for blood plateletpheresis at the Edgardo Rebagliati Martins-EsSalud National Hospital, Lima-Peru 2022, using a database of donors who came to donate platelets through plateletpheresis in the period 2015-2022. A descriptive, retrospective, and non-experimental research design was used. To evaluate the predictive models, the Python programming language was used using Google Colab, where the stages that involve the construction of a Machine Learning model were carried out, finding that the decision tree was the model with the best performance both when using unbalanced data (Precision=0.89; F1-Score=0.91; AUC=0.90) as balanced data using SMOTE (Precision=0.87; F1- Score=0.89; AUC=0.90), having a better predictive capacity in the classification of platelet donors versus other models.