Libraries at University of Nebraska-Lincoln
Date of this Version
Summer 4-8-2019
Abstract
Soft computing paves way many applications including medical informatics. Decision support system has gained a major attention that will aid medical practitioners to diagnose diseases. Diabetes mellitus is hereditary disease that might result in major heart disease. This research work aims to propose a soft computing mechanism named Improved Evolutionary Support Vector Machine classifier for CAHD risk prediction among diabetes patients. The attribute selection mechanism is attempted to build with the classifier in order to reduce the misclassification error rate of the conventional support vector machine classifier. Radial basis kernel function is employed in IESVM. IESVM classifier is evaluated through the performance metrics namely sensitivity, specificity, prediction accuracy and Matthews correlation coefficient (MCC) and also compared with existing work and our earlier proposed works.
Included in
Artificial Intelligence and Robotics Commons, Library and Information Science Commons, Theory and Algorithms Commons