Predicting the Presence of Poly Cystic Ovarian Syndrome using Classification Techniques

Abstract

PCOS is an endocrine disorder which occurs due to hormone imbalance. PCOS may leads to infertility, diabetes mellitus and cardiovascular diseases. It may be identified by physical appearance, ultrasound scanning and clinical trials. The PCOS ovary can be identified as the follicles which are arranged peripherally and measuring 2-9mm of size. The dataset used in this paper consists of 119 samples with 17 features which represents the physical appearance and psychological characteristics such as stress, exercising methods, eating habits, etc. The classification algorithms can be applied on these data to predict the present of PCOS. The aim of the paper is to compare the accuracy of the classification model and find the algorithm which best suites for the dataset in predicting the occurrence of PCOS