As a new version of support vector machine (SVM), least squares SVM (LS-SVM) involves equality instead of inequality constraints and works with a least squares cost function. A well-known drawback in the LSSVM applications ...