Neural Networks In The Stock Market Prediction

Forecast is necessarily an important link in scientific management and premise before policy-making and layout. It is necessary to forecast and analyze evolution trend of some system. One of the current forecast methods is time series forecast which constructs models according to the historical data before using it to forecast the future. Artificial neural network is an embranchment of artificial intelligent, originated in 1940s and is widely applied to many fields now. Neural network can study and reserve prevenient information and knowledge, which is the theoretical basis when used to forecast the future. As for nonlinear time series forecast, neural network is more efficient and precise than mathematical models.Artificial Neural network is a nonlinear dynamic system, which can realize the reflection of nonlinear relations among constants within the range of any accuracy, capable of solving nonlinear problems, learning network technology and integrating systems. It is able to satisfy the needs for the prediction of economy index, enabling the system to solve the nonlinear and unfixed problems, and hence improving the accuracy of prediction. In this paper, under the guidance of economic theory, we analyze the characteristics of economy system, based on the nonlinear and unfixed features of economy system, the theory on artificial neural network is applied to construct the modeling system of economy prediction. Then the modeling method to combine the multi evolutionary computation and Neural Network has been put forwords, which improves the accuracy of the system prediction.In this paper, definition, background, significance, several research situation of forecast and some usual forecast methods were commented in chapter 1, also five error targets in order to evaluate forecast precision were introduced in chapter 1. Chapter 2 summarized basic structures, algorithms and some BP networks existing problems, such as convergence rate, the global convergence and generalization. In chapter 3, instances were used to research optimized by BP neural networks. In chapter 4, experience data were worked by neural networks based on Genetic Algorithm, and the results were compared. Also, in chapter 5, the neural networks based on Particle Swarm Optimization algorithm forecast the data and the results were recorded and compared.Predecessors' research fruit is summed up in this paper, and proposes the concept and the opinion of writer self aimed at the defect, and throwing them into the practice, and does one's best to supply valid measure for stock market forecasting based on Artificial Neural Networks, and making great efforts in the interest of spreading the Artificial Neural Network technique.