The aim of this study is: poverty villages in Guangxi are the objects of this study, guided by varity theories and analyzed and evaluated systematically. On this bases, analyse the elements of poverty and classify the poverty villages of Guangxi according to the factors and their weighs which lead to the poverty. The purpose of classify the poverty villages is to made the celection and designment measures of poverty eradication more directedly and operatefully.This paper selects samples casually from more than 400 poverty eradication project villages in Guangxi ,draws out the elements of poverty, and then select and design a sets of system of indicate of evaluation .Based on fully investigate, quantify the system of indicate of evaluation according to ideal of experts and villagers ;designs a model of artificial neural network to classify the poverty villages,and write the classify procedure with MATLAB 7.0,make classify softwire; then input the quantited data and the ideal of experts to the model to sdudy and simulate, we can classify the poverty villages to three types: type of restricted by environments poverty village,type of restricted by personal abilities poverty village, and type of restricted by rights poverty village, thus realizes the classification of poverty village.The result of using this modle to classify the poverty villages is more accurate and easy than the expert classification, it proves that the model of artificial neural network to classify the poverty villages designed in this paper is advanced in technology, being practical useful and operateful in practice, and it has a renew significance for theory and an important reference value in the study of complicated social system.
|