Thyroid Disease Prediction Using Machine Learning Techniques

Thyroid disorders are primarily observed in people in the age range of 30 to 60 years; however, the disease is said to be very prevalent in women. Some of the risk factors for thyroid disorders include other diseases that are autoimmune in nature, a family history of thyroid disease, and past exposure to radiation of the neck. More correctly, thyroid diseases are meaningful metabolic disorders and determinants of health. Toward this end, this study used a large dataset from the patient medical record to predict the onset of thyroid disease. The models that are used in this work are K-Nearest Neighbors (KNN), Multi-Layer Perceptron (MLP), Support Vector Machine (SVM), Logistic Regression, and Naive Bayes. Further each of the models’ performances is also assessed based on recall, accuracy, and precision. The MLP model developed a good predictive ability in comparison with the other models, along with an accuracy of 99.4 percent. Therefore, this study puts forward a very important message concerning the vital importance of machine learning.