An intelligence model for detection of PCOS based on k‐means coupled with LS‐SVM

Abstract

Polycystic ovary syndrome (PCOS) is a hormonal disorder that affects women at an early age. Manual detection of PCOS is a challenging task for specialists, however, detection of PCOS as quick and accurate could save the lives of millions of women over the world. Current studies use high dimension features which leads to a low estimation accuracy, and high execution time. However, in this article, we develop a new intelligence system to classify PCOS based on k‐means coupled with a LS‐SVM (K‐M‐SVM) using a lower number of features. The original dataset is preprocessed and then k‐means is applied to select the most powerful features based on Euclidean distance to classify PCOS. It was found that the k‐means cluster had a high potential in selection the most influential features and eliminating the poor ones. As a result, a total of six features are chosen to represent PCOS data from the original features. The selected feature set are fed to the LS‐SVM to classify them into healthy and no healthy segments. Our findings showed that the proposed model (K‐M‐SVM) outperformed the state of the art, and it gained an accuracy of 99%.