Study On The Predictive System Of The Welfare Lottery Based On BP Network

With the popularity of the welfare lottery and soccer lottery,it is increasing its investment enthusiasm。The people who invest the welfare lottery have to obtain a return on the basis of their own experience using some forecasting methods,while some experts have given their views to forecast the winning numbers for the next. But the interference problem and nonlinear problems have not been properly soluted among these forecasting methods for the lottery system. With the development of artificial intelligence technology, artificial neural network theory has got fast development too, Back Propagation neural network is one of the most representative neural network model among the various kinds of neural networks, and has found extensive application. This paper applied the theory to forecast the lottery.This paper introduced BP neural network and BP algorithm detailedly, analyzed the main defects and the reasons produced by this algorithm and put forward the corresponding improving measure,for instance:momentum item,adjusting the learning rate.First, according to the rules of the game,made sure the neural network structure of input and output layer, designed BP network at three aspects: Whether the implicit layer and the output layer uses the activate function or not;Implicit layer number;The best piece of the implicit layer node. Pass the contrast toward the study and the estimate result analysis, get to draw a conclusion:The implicit layer and the output layer used activate function differently;Choose three layer BP networks;The node of the implicit layer counts to 17, making sure BP network structure thus.Using the trained a good network to carry on a test, the correctness attains 76%, comparing other estimate softwares of the accuracy raise 2~300%. On this basis, this research will be designed as a predictive software, and gave the software architecture, function and the interface which is easy to operate . For the prediction of welfare lottery the study opened up a new avenue for both theoretical guidance and immediate value of the application.