基于差值灰色RBF网络模型股票价格预测研究

Stock price prediction based on Difference Grey RBF neural network model

针对RBF神经网络的预测精度受样本数据随机性影响较大,而灰色理论能弱化数据随机性的特点,提出了差值结合法将灰色GM(1,1)模型和RBF神经网络模型有效地结合起来,构建了差值灰色RBF网络预测模型。并利用此模型进行股票价格预测,实证结果表明:该模型预测稳定性较好,预测精度高,平均预测误差为0.68%,与BP神经网络和RBF神经网络相比具有更好的泛化能力和更高的预测精度,在股票预测中具有一定的实用价值。

To solve the problem of the large variation in prediction accuracy of RBF neural network that is influenced by random data is proposed, a difference combination method, which combines gray GM （1,1） model and RBF neural network model was developed to establishes the Difference Gray RBF model. This model was used in stock prediction. The results show that the model can pre- dict stock price stably and accurately, with an average error about 0. 68%. Compared to BP neural network and RBF neural network, the Difference Gray RBF has a better adaptability and higher ac- curacy of prediction, which is of some value in stock prediction.