股市价格趋势预测研究

Stock Price Trend and Detail Multiple Prediction

针对直接对价格序列进行预测偏差较大的问题,提出了在处理有趋势特征的价ANFIS(Adaptive-Network-based Fuzzy Inference System)格数据中,用数字滤波器获取价格趋势和围绕趋势波动的变化细节特征,对两者分别采用和模型进行预测。此方法应用于股票价格ARANFIS序列预测中,收到较好效果。

In order to overcome the shortage of ANFIS(Adaptive-Network-based Fuzzy Inference System) which have notable error when applied in prediction of price series, an approach was proposed that we get the price trend and the detail fluctuation on the trend with digital filter and predict them with AR model and ANFIS model separately. The prediction results of stock price with the approach approves that the accuracy is improved notably. ;;;