Title :
A Novel Method of Pattern Recognition for Honey Source Based on Visible/Near Infrared Spectroscopy: Genetic Algorithm Combined with Support Vector Machine
Author :
Yang, Yan ; Nie, Peng-Cheng ; Zhang, Wei ; He, Yong
Author_Institution :
Coll. of Bio-Syst. Eng. & Food Sci., Zhejiang Univ., Hangzhou, China
Abstract :
Visible/near infrared spectroscopy (Vis/NIRS) appears to be a rapid and convenient nondestructive technique that can realize the qualitative analysis and quantitative analysis for many agriculture products. In this study, a novel non-destructive pattern recognition method for honey source was developed base on Visible/Near infrared Spectroscopy. The four types of honey, linden, Chinese milk vetch, locust and Wild Chrysanthemum were analyzed. The sample sets consists of 200 samples for calibration set and 32 samples for predict set. The SIMCA, PCA-SVM and GA-SVM classification algorithm were employed to build the discrimination model respectively. The accuracy of discrimination model was used to judge the discrimination of model. In order to simplify the model, the significant wavelengths were extracted by GA algorithm and be used to construct discrimination model base on SVM. Finally the discrimination of three models was compared respectively. The result indicated the GA combined with SVM algorithm offer a new approach to recognition for honey source.
Keywords :
agricultural products; genetic algorithms; infrared spectroscopy; pattern classification; support vector machines; Chinese milk vetch honey; GA-SVM; PCA-SVM; SIMCA; agriculture products; classification algorithm; genetic algorithm; honey source; linden honey; locust honey; nondestructive technique; pattern recognition; support vector machine; visible-near infrared spectroscopy; wild chrysanthemum honey; Classification algorithms; Educational institutions; Gallium; Genetic algorithms; Principal component analysis; Spectroscopy; Support vector machines; Honey; Pattern recognition Introduction; Principal component analysis (PCA); Soft Independent Modeling of Class Analogy(SIMCA); Visible/ near infrared spectroscopy (Vis/ NIRS); genetic algorithm (GA); support vector machine (SVM);
Conference_Titel :
Artificial Intelligence and Computational Intelligence (AICI), 2010 International Conference on
Conference_Location :
Sanya
Print_ISBN :
978-1-4244-8432-4
DOI :
10.1109/AICI.2010.114
