Polycystic ovarian syndrome (PCOS) stands as a prevalent endocrinological disorder significantly impacting female fertility and health. Despite its prevalence, the exact etiology of PCOS remains elusive. Key symptoms include weight gain, hirsutism (excessive facial hair growth), acne, alopecia (hair loss), skin pigmentation changes, and irregular menstrual cycles, all of which pose significant hurdles to achieving pregnancy. Early detection of PCOS can reduce complexity. Thus, in order to reduce difficulties, an accurate and timely PCOS screening system is necessary. Machine learning (ML) is one of the detection strategies that perform the best because to its feature extraction capacity. It has shown encouraging results in a number of research. PCOS has been predicted and diagnosed using a variety of machine learning methods, including XGBoost, Random Forest, Logistic Regression, decision trees, and Support Vector Machines This study intends to draw attention to the researchers by providing a contextualized and descriptive summary of all the technologies currently in use for ML algorithms-based PCOS detection.