Empirical Analysis of Machine Learning Algorithms for Predicting Thyroidism

This paper focuses on early-stage prediction of Thyroidism, particularly among young women, due to sedentary and stressful modern lifestyles. Various machine learning techniques (decision trees, random forests, support vector machines, naive bayes, and K-nearest neighbors) were simulated for disease probability assessment. Thyroid disorders result from hormonal imbalances caused by stress or infections. The study identifies key clinical indicators and employs K-Nearest Neighbor (KNN) as the best-performing classifier for thyroid prediction. This research aims to improve thyroid disease diagnosis using effective classifiers and machine learning algorithms based on data classification accuracy and performance evaluation.