Implementation of Various Machine Learning Algorithms to Predict Polycystic Ovary Syndrome

Abstract

Polycystic Ovary Syndrome (PCOS) is a hormonal condition that affects women of reproductive age and can cause acne, facial hair growth, hair loss, infertility, irregular menstrual cycles, and weight gain. Early detection and treatment of PCOS can be challenging. We propose a system that uses machine learning algorithms to predict and diagnose PCOS using minimal parameters. We used a dataset from the open-source database "KAGGLE" and identified the top 10 to 15 features after speaking to gynaecologists. Four machine learning algorithms were used to train, validate and test the model, including the Random Forest classifier, logistic regression, Decision tree classifier and Chi-Square algorithm. Our results show that the Random forest classifier (Chi-Square) has the highest accuracy compared to the other algorithms. Our system can provide early detection, prognosis, and treatment suggestions for PCOS, which can improve the quality of life for women affected by this disorder.