基于小波神经网络的上证50指数多步预测

以神经网络为代表的人工智能模型对股票价格具有良好的预测效果,但是该智能模型侧重于单步预测,很难满足实际股票预测的要求。提出基于小波和神经网络相结合的股票指数多步预测智能模型。选取上证50指数为建模数据,运用小波分解将上证50指数收盘价序列分解成不同尺度的分层数据，依据迭代策略,利用BP神经网络分别预测小波分解后的各层数据，最后将各层的预测结果使用小波重构成原始股票收盘价的预测数值。结果表明，基于小波神经网络的多步预测模型具有良好的多步预测效果。

Multistep Forecasting SSE 50 Index Based on Wavelet Neural Network

Represented by the neural network model of artificial intelligence to the stock price has good prediction effect, but the intelligent model focuses on the single-step prediction, it is difficult to meet the requirements of actual stock prediction. In this paper, based on the combination of wavelet and neural network put forward multi-step predictive model of stock index .Select the Shanghai 50 index closing price as the modeling data, Wavelet decomposition, the Shanghai 50 index closing sequence decomposed into hierarchical data at different scales, according to an iterative strategy, the use of layers of BP neural network prediction data were wavelet decomposition, and finally the layers prediction wavelet reconstructed original stock predictive value closing price. The results show that the multi-step prediction model based on wavelet neural network has good multi-step prediction.