|International Journal of Applied Information Systems|
|Foundation of Computer Science (FCS), NY, USA|
|Volume 12 - Number 41|
|Year of Publication: 2023|
|Authors: Terlumun Emmanuel Togor, Joshua Abah, Dekera Kenneth Kwaghtyo|
Terlumun Emmanuel Togor, Joshua Abah, Dekera Kenneth Kwaghtyo . Development of Thyroid Disease Prediction Model in Nigeria. International Journal of Applied Information Systems. 12, 41 ( Sep 2023), 33-47. DOI=10.5120/ijais2023451950
Thyroid Diseases are often diagnosed in Nigeria's endocrinology clinics and are reported as the second most common endocrine disease encountered amongst Nigerians, especially women. This led to a vast volume of thyroid data. When applied to machine learning algorithms, the data can greatly benefit endocrinologists, patients, and health organizations. However, diverse machine-learning methods have been in place for thyroid disease prediction leveraging the UCI thyroid data for experimental purposes. Conversely, this study differs by utilizing indigenous thyroid data to predict the thyroid conditions: hypothyroidism, hyperthyroidism and euthyroidism. The dataset was pre-processed using pandas and NumPy libraries. Random Forest classifier and Support Vector Machine classifier were trained using the indigenous dataset. Experimentally, the model was evaluated using Accuracy, Precision, F1-measure, Sensitivity and the Receiver Operating Characteristic (ROC) curve – Area Under the Curve AUC. The classification results show that the Random Forest classifier obtained the best accuracy of 99.30%, while the Support Vector Machine classifier achieved an accuracy of 98.60%. The study achieved the goal of turning the data comprising of thyroid conditions gathered from the Federal Medical Centre, Yenagoa, Bayelsa State, Nigeria into a powerful tool to support endocrinologists, patients, and health organizations.