The optimized support vector machine with correlative features for classification of natural spearmi

The performance of support vector machine (SVM) hybridized with two other methods for classification of chemical patterns was investigated. It was found that SVM for classification can be sensitive to noise and be affected by multicollinearity between attributes similar to other methods such as multivariable analysis and neural networks. The kernel function, its parameter and penalty factor C are the main factors affecting the classification performance of SVM. Correlative component analysis (CCA) was used to eliminate multicollinearity and noise of original sample data before classified by SVM. To improve the classification performance of SVM and obtain the optimal discriminate function, Eugenic Genetic Algorithm (EGA) was used to optimize the parameters of SVM. Finally, a typical example consisting of two classes of natural spearmint essence was employed to verify the effectiveness of the new hybridized approaches including CCA-EGA-SVM. The classification accuracy of this new method is much better than that obtained by SVM, CCA-SVM, CCA-SOM, and GA-CG-SVM.