Comparative Analysis of Machine Learning Algorithms in Thyroid Disease Prediction

Thyroid disease is estimated to be the most common illness in the twenty-first century. The thyroid hormones are produced by a butterfly-shaped gland in the throat. The crucial function these hormones play in bodily metabolism, mental health, and general well-being. In this study, we present a method for accurately analysing thyroid function in an unbalanced dataset using Machine Learning (ML) algorithms. This dataset's accuracy score is ambiguous because most target variable classes are 1. Performance measurements are accuracy and recall, but when class distribution is asymmetrical, the F1-score balances the other parameters. The performance is evaluated using the F1-score, which is one of the best output metrics for unbalanced classification tasks. Simulation results shows that the Decision Tree (DT) algorithm outperforms the other algorithms with an F1-score of 0.9957, the Area Under Curve (AUC) score of 0.9917 and the kappa coefficient of 0.978 in the disease prediction.