Research Based On Data Mining In Forecasting Securities

View of the problems that exist in forecasting of current trend of the domestic stock market, such as: the huge amount of computation process forecasts, projections and other unsatisfactory results, the use of data mining in this paper several methods used to establish prediction model, predicting the outcome of the process analysis and prediction, in order to seek the point of data Mining stocks forecast algorithm and the integration.It is use the time-series methods, Markov chain model, the artificial neural network model in this paper. Applications of time-series forecasting methods have been used in a simple moving average forecast, a forecast of exponential smoothing, two exponential smoothing prediction method. time-series methods have the most simple in forc- easting trend of the stock, but have the largest error. Application of Artificial Neural Network, BP algorithm and used genetic algorithms to train the network weights, but also adopted the learning rate and momentum method of combining adaptive adjustment of the strategy, the price of the securities markets, modeling and forecasting. Artificial neural network model in spite of the projected best effect, but the volume calculated. Compared to the first two methods, Markov chain model is more suited to forecast the movement of the stock, the smaller amount of computation, the effects of the artificial neural network model forecast similar results with improve forecasting methods.