Observation on Polynomial and Radial Basis Function in Support Vector Machine (SVM) Tuned parameters for Agarwood Oil Quality Grading

Mohamad Azim Wafiq Khairuddin, Nurlaila Ismail, Nor Azah Mohd Ali, Mohd Hezri Fazalul Rahiman,Saiful Nizam Tajuddin, Mohd Nasir Taib

Abstract

Agarwood oil is well-known as exorbitant lubricant, extricated from the adhesive of aromatic heartwood. The oil is profoundly request in the market particularly from Asia, Turkey, Iran and India in view of its one of a kind scent. First-class agarwood oil is one of the most overpriced common organic materials in the nature. For this on-going research in classifying quality of the agarwood oil, the observation on Polynomial and Radial Basis Function (RBF) in Support Vector Machine (SVM) tuned Parameters for agarwood oil quality grading. The work involved of feeding abundances (%) as input to SVM network training and quality of the oil as output. The Polynomial and RBF kernels were used during training stage. After that, the development of SVM model was tested using testing dataset. The complete analysis work was achieved using MATLAB R2015a version. The outcome of the finding proved that Polynomial were better than RBF in classifying the agarwood oil. The research advantages to future work and also the utilization for agarwood oil investigate region particularly its quality grouping.