@article {10.3844/jcssp.2017.590.599,
article_type = {journal},
title = {Determination of SVM-RBF Kernel Space Parameter to Optimize Accuracy Value of Indonesian Batik Images Classification},
author = {Budiman, Fikri and Suhendra, Adang and Agushinta , Dewi and Tarigan, Avinanta},
volume = {13},
number = {11},
year = {2017},
month = {Oct},
pages = {590-599},
doi = {10.3844/jcssp.2017.590.599},
url = {https://thescipub.com/abstract/jcssp.2017.590.599},
abstract = {Image retrieval using Support Vector Machine (SVM) classification verydepends on kernel function and parameter. Kernel function used by dot productsubstitution from old dimension feature to new dimension depends on imagedataset condition. In this research, parameter of Gaussian /Radial BasisFunction (RBF) kernel function is optimized using multi class non-linear SVMmethod and implemented to training and test datasets of traditional Indonesian batik images. The batik images dataset is limited to four geometric motifs textures,which are ceplok/ceplokan, kawung, nitikand parang/lerang. Discrete WaveletTransform level 3 daubechies 2 is used to result feature dataset of traditionalbatik images dataset of four classesgeometric motifs textures. The batikimages are used for training and test dataset in SVM-RBF kernel parameteroptimation to maximize accuracy value in non-linear multi-class classification.Cross Validation and Grid-search methods are used to analyze and evaluateSVM-RBF kernel parameter optimation. Confusion matrix measurement method isused to result accuracy value in every evaluationconducted in every combination of cost function/C and gamma/γ as SVM-RBF kernel parameter. Maximumaccuracy parameter value is C = 27 and γ = 2-15 achieved by 10 times evaluation wit different testdataset for each evaluation. Maximum accuracy value is 0.77 to 0.86.},
journal = {Journal of Computer Science},
publisher = {Science Publications}
}