Study of Flue-Cured Tobacco Classification Model Based on the PSO-SVM

摘要

In this study, we study the flue-cured tobacco classification model based on the PSO-SVM. Firstly we use the Gaussian Radial Basis Function (RBF) as the kernel function of SVM and then use the Particle Swarm Optimization algorithm (PSO) to optimize the structural parameters of the SVM classifier, established the flue-cured tobacco classification model based on the PSO-SVM. Collecting a wide range of tobacco data in Qujing Yunnan Province, to train and validate the model. At last, compared with the grid parameter optimization and genetic algorithm-based parameter optimization model, the results show that the proposed model based on particle swarm optimization with high prediction accuracy and better adaptability when used in tobacco grading.