Classification of Polycystic Ovary Syndrome Based on Correlation Weight Using Machine Learning

Abstract

Polycystic ovarian syndrome (PCOS) is the most common endocrine pathology in reproductive-age women worldwide. Research has shown that the application of machine learning (ML) and data mining (DM) can have a positive impact in this condition's diagnosis. This study aims to develop a model to identify patients with PCOS using different scenarios based on correlation weights. Five DM techniques were applied, namely random forest (RF), decision tree (DT), naive bayes (NB), logistic regression (LR), and artificial neural network (ANN), to determine the best model, which was the RF classifier. Additionally, the results show that the model was able to predict PCOS with 93.06% of accuracy, 92.66% of precision, 93.52% of sensitivity, and 92.59% of specificity. Compared with a previous work conducted by the authors, the feature selection-based solo on the correlation weight decreased the accuracy values by 1.9%, precision by 3.7%, sensitivity by 0.3%, and specificity by 3.6%.