Investigation of the soybeans disease motivates the need for a programmed detection system. Automated detection using a vision system and pattern recognition are implemented to detect the symptoms of nutrient diseases and also to classify the disease group. Research before the show that disease recognizing can be conducted with a classification such as Suppor Vector Machine. Reminding, one of the advantages of Support Vector Machine, is able to increase performance on generalization with choosing the exact kernel function, thus on this research would like to find out which kernel function appropriate to the classification problem on soybeans disease using two kinds of the kernel function, Radial Basis Function (RBF) and Linear. Based on the performance result conducted with soybeans dataset, both of them can work well on a classification problem. However, from both function kernel, Radial Basis Function (RBF) classify better than the other with an accuracy of 83% of correct classification
Copyrights © 2019