Self-Organizing Map Analysis on Peanut Yield and Agronomy Characteristics

The model system between peanut yield and agronomy characteristics which is nonlinear, irreversible and dissipative. The objective in the study was the peanut cultivated in the different ecological regions in Shandong province, aimed to establish the new non-nonlinear model based on Self-Organizing Maps (SOM) to improve the cultivation information of peanut growth process. In the article, applying SOM network achieved the cluster between peanut yield and agronomy characteristics about 4 variables, involved in plant height, branches, full pods and peanut yield ratio. MATLAB 7 software is used to classify 60 samplings of peanut yield and agronomy characteristics. It is concluded that the SOM network can respond the complicated information classification among each peanut yield, during the analysis, the results also showed SOM method is the most suitable for peanut yield and characteristics classification, especially analysis of clusters on basis of peanut agronomy parameters,so the study can be applied on agronomy characteristics and peanut yield of the different ecological regions in Shandong province.