The Application Research Of Data Mining Algorithm On Stock Market Prediction

The stock market, which is the main character of market economy, attracts millions of investors from its birth. Because high risk high payoff is the character of the stock market, Investors care for stock market, analysis stock market and try to predict the trend of stock market. But the stock market is influenced by political and economic factors, the internal discipline is very profound, so the prediction results of traditional prediction technology are unsatisfied.Facing the existing problems in stock market prediction, such as the nonlinear property and high noise property of stock price (stock index), the paper try to find the joint point of data mining algorithm and stock prediction by studying the application of data mining algorithm in stock prediction.Firstly, the paper introduces all kinds of data mining algorithm and the research situation of data mining algorithm, puts forward to using data mining algorithm in stock prediction in order to predict stock market more accurately.Secondly, the paper uses data mining algorithm to predict stock index, adopts the Grey System model and Artificial Neural Network model to predict stock index separately. The prediction results show that in the short-term prediction of stock index, Grey System model is an effective model, and Artificial Neural Network model also has good prediction results . Because the prediction results of single model is still unsatisfied , the paper put forward the combined grey neural network model and use it to predict the stock market index, the prediction results is better than single model's results.Lastly, the paper uses data mining algorithm to predict the individual stock price. The paper puts the common used stock market technical index into the Artificial Neural Network model, uses BP network and Radial Basis Function Network to predict the individual stock price separately. At the same time ,the paper adopts exponent smoothness method in the same question, and compares the prediction results of different methods. The numerical experiments show that the predictionresults of Radial Basis Function Network are better than BP network's, the prediction results of BP network are better than exponent smoothness method's.