Optimized Thyroid Disease Classification Using Nature-Inspired Algorithms: GWO & WOA

Recently, there has been an upsurge in the number of cases of thyroid disease. Thyroid function is essential for metabolism, making the early diagnosis of thyroid dysfunction an urgent matter. The issue of class imbalance has not been thoroughly examined, even though there are multiple publications on the topic of thyroid disease detection. Furthermore, the binary-class problem has been the primary emphasis of previous research. This study intends to address these concerns by using the suggested strategy, which takes into account ten distinct thyroid illnesses. In order to choose the best features from the Thyroid dataset, this research proposes two optimization algorithms: Grey Wolf and Whale. The features chosen by GWO are five, while those chosen by WOA are eight. GWO achieves an accuracy of 98.6% and WOA of 98.11% when it comes to selecting the ideal reduced set of features for classifying the thyroid disease.