Comparative Analysis of Classification Methods for PCOD Prediction

Today's women often experience anovulation, infertility, and preterm abortions. The condition polycystic ovarian syndrome (PCOD), which affects women of reproductive age, has been strongly associated with infertility. This hormonal disorder affects females who are in their reproductive years. Menstrual cycles in women that are late or nonexistent are the result of hormonal imbalance. Some women may be infertile because of polycystic ovarian syndrome (PCOS), which can also present as extreme weight gain, abnormal hair, eczema, alopecia, hyperpigmentation, trouble becoming pregnant, and hormonal abnormalities. Once the illness has been discovered, there is no cure, although the right care may lessen the symptoms. PCOD is very challenging to diagnose because of the vast spectrum of symptoms and the occurrence of other related gynecological issues. Patients with PCOD face significant challenges due to the expensive nature and extended commitment of many clinical tests. We suggest a method for addressing this problem based on a complete and vital set of features that could help in the timely identification and prediction of PCOD treatment. Random Forest, Support Vector Machine, Logistic Regression, Gaussian Naive Bayes, and K-Nearest Neighbors were among the classifiers we employed in machine learning to predict whether or not a woman would develop PCOD.