神经元网络在股价预测中的应用

利用时延神经元网络模型（ＴｉｍｅＤｅｌａｙＮｅｕｒａｌＮｅｔｗｏｒｋ）对四川长虹的股价作了预测。股价的涨跌预报可视作高维空间的非线性分类问题，本文使用增益可调的反向传播算法，对其走势作了预报。借助前馈神经网络对非线性函数的逼近能力，本文对四川长虹股价这个时间序列作了连续若干天的一步预测。最后，我们采用不同形式的误差函数对预测结果作了比较。

The Application of Neural Networks to Stock Price Forecasting

This paper presents the application of time delay neural networks to the prediction of the Sichuan Changhong Stock Price. The rise/fall prediction of the stock price can be regarded as the classification problem in the highdimension space. Using the back propagation learning algorithm with an adaptable slope of the activation functions, we report on the rise/fall prediction of the stock price. The nonlinear functional approximation capabilities of feedforward neural networks lead us to forecast time series, and we have undertaken the one step stock price forecast. In the end, different error functions are compared with respect to their performance on prediction.