RETRACTED ARTICLE: Optimal feature-based multi-kernel SVM approach for thyroid disease classification

Abstract

Thyroid diseases are across the board around the world. In India as well, there is a critical issue caused because of this disease. Different research studies estimate that around 42 million individuals in India suffer from the ill effects of “thyroid diseases.” Diagnosis of health situations is an energetic and testing undertaking in the field of therapeutic science. Our proposed model is to classify this thyroid data utilizing optimal feature selection and kernel-based classifier process. We will create classifications models and its group show for classification of data using “multi kernel support vector machine.” The novelty and objective of this proposed model as feature selection, it’s used to enhance the performance of classifying process with the help of improved gray wolf optimization. Reason for this optimal feature selection as the insignificant features from unique dataset and computationally increment the performance of the model. The proposed thyroid classification results in the accuracy, sensitivity, and specificity of 97.49, 99.05 and 94.5%, compared to the existing model. This performance measure is computed from the confusion matrix with the diverse measures contrasted with individual models and in addition to the existing classifier and optimization techniques.