Predictive Analysis for Thyroid Diseases Diagnosis Using Machine Learning

Thyroid disease is a condition in which the thyroid gland does not produce enough hormones. The symptoms of thyroid disease vary depending on the type (hypothyroidism, hyperthyroidism, or other). Generally sleeping trouble, anxiety, losing weight, fatigue, gaining weight, forgetfulness, and many other complexities are caused by hyper and hypothyroidism. So, diagnosing thyroid diseases is a vital issue as the diseases can be cured through proper and timely diagnosis. Recently machine learning techniques are used for predicting thyroid diseases. In this study, we have proposed a thyroid diseases prediction model through three machine learning classification algorithms namely K-Nearest Neighbor (KNN), Naive Bayes, and Decision Trees. Experiments are performed on thyroid data of the UCI machine learning repository. The dataset has three classes named normal, hypothyroid, and hyperthyroid. Through 10-fold cross-validation, the performances of the three algorithms are tested on several parameters such as Accuracy, Precision, F-Measure, and Recall. The decision tree was the most accurate, with a 99.7% accuracy rate over Naïve Bayes and KNN in the three-class thyroid diseases classification problem.