A Novel Feature Selection Method for Classification of Medical Data Using Filters, Wrappers, and Embedded Approaches

Feature selection is the process of identifying most relevant features from given data having a large feature space. Microarray datasets are comprised high-quality and very few samples data. performed on such to identify optimal subset. The major goal improve accuracy by minimal For this purpose, proposed research focused analyzing effective algorithms. A novel framework which utilizes different methods filters, wrappers, embedded Furthermore, classification then selected classify using support vector machine (SVM) classifier. Two publically available benchmark used, i.e., dataset Cleveland Heart Disease dataset, for experimentation analysis, they archived UCI repository. performance SVM analyzed accuracy, sensitivity, specificity, f-measure. 94.45% 91% achieved each respectively.