Based On Neural Network Financial Network Location System

Selection for financial locataion evaluation is the selecting process for financial structure that is under some economical and social condition. It is very important to select proper location for financial facilities, especially facing to the fierce market competition and an increasingly complex economic environment. Howervr, the traditional methods for location selection reveal some shortcoming: there is obvious difference between the abstract mathematical models and the practicalities; it is very hard to consider all the complex and abstract factors with the traditional methods; the abundant data can not be organized and analyzed efficiently; the traditional methods can not provide the decision-markers some visual and interactive tools to analyze the location problems.With the development of Geographical Information System (GIS), especially Spatial Decision Support System, which integrates GIS and modeling technology, is introduced to solve location evaluation problems, there is a new way to solve the difficulties baffling the traditional methods.Based on GIS platform, we introduce the theories on financial location evaluation in detail, including its definiation, principle, target and general thinking. Then we raise the financial location evaluation model by using Artificial Neural Networks. The model combines Genetic Algorithm and BP network which implements the study process of BP network by using Genetic Algorithm. This method improves BP network's learning rate and covers the disadvantages of Genetic Algorithm. With the analysis of Neural Networks, especially the Genetic Algorithm and BP networks, we improve the Genetic Algorithm, mainly on its parameters' (the number of groups, the rate of crosspoint and the rate of variation) selection, determination and adding the BP factor etc. By integering the BP networks' parameters, we can create the genetic code. Then we can use our improved Genetic Algorithm to determine the BP networks' ultimately parameter' value. Finally, using the BP arameter, we can forcast and evaluate the candidate location.At last, we implement the location evaluation system, which is based on GIS. We also introduce this system's framework and functions in general.