Analysis On Economics Differentiations Of Counties Based On BP Neural Networks In Lin Fen City

Regional economic is comprehensive application system whichstudy of different regional economic development, spatialorganization and their relationship. A regional economicdevelopment process and characteristics of internal change isaffected by many factors, such as income, consumption, investment,urbanization and other factors of the influence and restriction. Onthe one hand, because of the influence of various factors on thedegree of the interference of the whole system is different, so eachregional economic system’s dynamic and complicated nonlinearsystem; At the same time,because of the their own uncertainty ofthe various influencing factors,combined with external interferencefactors, makes the trend of regional economic system in the originalcondition can present certain fluctuation, the regional economicforecast to bring very great difficulty. On the other hand,of countyterritory economy development level of evaluation, the domesticmany scholars have carried out extensive research, most are basedon different statistical methods, and evaluated for model building.But from an existing application, most of these methods or modelsfor linear methods and models, there are different characteristicsand application limitation,it is difficult to grasp the regional economy of this kind of complex nonlinear system reasonableevaluation. Selecting which influence factors and setting up whatkind of evaluation model and how to review the evaluation resultsis an important content of this study.Application of ANN theory and method, adopting8economicindexes, based on the BP neural network toolbox, the authorsdeveloped BP neural network model with8neurons in input layer,5neurons in hidden layer and1neuron in output layer in economicdevelopment level evaluation, with which Lin fen economicdevelopment level in2011was evaluated, and divided all the17counties into4types. Moreover, MapInfo software is applied tomake a thematic map to express the Economic development leveldisparities of these seventeen counties. Besides, the paperanalyzes frequency distribution features and calculates variationcoefficient, weighted variation coefficient, William coefficient andmaximal and minimal coefficient, finding:①economiccomprehensive evaluation value "One axis" as the drive, drive the"two belts" coordinated development;②its frequency distributionhas positive skewness features, and a slightly bigger proportion ofcounties in the third degree and fourth degree than the first degreeand second degree;③internal differentiations of three regionsincrease from "One axis" to "two belts".