数据挖掘在股票价格组合预测中的应用

Application of Data Mining in Stock Price Prediction

研究股票价格变化预测问题,股票价格受多种影响,导致具有突变性、非线性和随机性,单一预测方法只能描述股票价格部分变化规律,预测精度低。为提高股票价格预测精度,提出一种基于数据挖掘技术的股票价格组合预测模型。根据股票价格变化特点,首先对其线性变化规律进行建模预测,并对非线性变化规律进行建模预测,最后将两种预测结果进行融合,得到股票价格的最终预测结果。仿真结果表明,相对于单一股票价格预测模型,组合预测模型提高了股票价格预测精度,降低了股票价格预测误差,更加全面、准确反映了股票价格的变化规律,是一种有效、高精度的股票价格预测参考手段。

Study the stock price changing trend. Stock prices are influenced by many kinds of effects, single pre- diction method can only describe the stock price change regula. In order to improve the precision of prediction of stock prices, the paper put forward a data mining technology based on combined forecasting model of stock prices. According to the characteristics of stock price changes, first, its linear variation prediction model was built, and then the nonlinear variation modeling was carted out. Finally, the two prediction results were fused to get the final forecas- ting result of stock prices. The simulation results show that, compared with the single stock price forecasting model, the combination forecasting model improves the stock price prediction accuracy, reduces the stock price prediction er- ror, and is more comprehensive and accurate.