Kernel Functions of SVM: A Comparison and Optimal Solution
Classification with better accuracy for all type of data set for a single classifiers is still a challenge in the domain of Machine learning. Improving the efficiency of classifiers is still a challenge for researchers. This notion has motivated to give comparative solution for the classifiers with the proposition of optimal solution. The classification algorithm can be applied for many application to improve its accuracy such as Gender and age classification, face etc. This is just to relinquish a boost to the svm algorithm researches in the various classification fields through the different kernel functions. The proposed methodology has propounded an optimal solution on the usage of kernel functions. There have been many researches on the kernel function comparisons. Here, a better and accurate solution has been applied on dataset for male female and transgender which is new gender has been classified and has been expectant to come up with a better accuracy.
KeywordsSupport vector machine Kernel function Classification
This chapter does not contain any studies with human participants or animals performed by any of the authors.
- 1.Hofmann, M.: Support vector machines—Kernels and the kernel trick. Notes, 26 2006Google Scholar
- 2.Howley, T., Madden, M.G.: The genetic evolution of kernels for support vector machine classifiers. In: 15th Irish Conference on Artificial Intelligence (2004)Google Scholar
- 3.Liu, L., Bo, S., Xing, W.: Research on Kernel Function of Support Vector MachineGoogle Scholar
- 4.Lin, H.-T., Lin, C.-J.: A study on sigmoid kernels for SVM and the training of non-PSD kernels by SMO-type methods. Submitt. Neural Comput. 3, 1–32 (2003)Google Scholar
- 5.Luss, R., d’Aspremont, A.: Support vector machine classification with indefinite kernels. In: Advances in Neural Information Processing Systems (2008)Google Scholar
- 10.Platt, J.: Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods. Adv. Large Margin Classif. 10(3), 61–74 (1999)Google Scholar
- 13.Baudat, G., Anouar, F.: Kernel-based methods and function approximation. In: 2001 Proceedings of the International Joint Conference on Neural Networks. IJCNN 2001, vol. 2. IEEE (2001)Google Scholar