A Study On Decision-Making Method Of Stock Based On Genetic Neural Network

Stock market is a highly complex non-linear dynamic system. The law has certain auto trendy. The law is not only restricted by the value law, but also influenced by political, economical, psychological and several extra elements at the same time. While conventional quantitative predictive method is based on the mathematical statistics, there are lots of difficulties faced of researching the stock market trend under so many uncertain elements.Focus on lots of problems during the process of stock market prediction. such as: the nonlinear character of stock market, the difficult of stock price predictions etc. this essay is aiming to give deeply research and explore the question which is supported by the stock operation, also designed a double ANN system and mentioned a stock decision support model. At first, the essay gave a simple description of the basic research method of adapted stock market prediction. And then gave the advantages and shortcomings of the analyzing method for existed all kinds of stocks. Secondly, the essay mentioned the concept of dynamic index, also used this index for analyzing and judging the rise and fall and knee point of stock. This avoids the prediction problems through predicting the stock price. Again mentioned double nerve network system, making judgment by using the judgment of network predicting results, increasing the accuracy and reliability. In final, the author builded up a decision support model based on nerve network, by aiming at the shortcomings of the model, the author brought out an improving method, which made this decision support model more adaptable.At last, the author used MATLAB7 to make demo prediction to above system and model and get lots of real data.