基于神经网络的黄金期货价格的预测

Based on neural network prediction of gold futures prices

黄金期货市场是一个极其复杂的非线性动力系统,由于神经网络具有很强的非线性逼近能力,理论上能无限逼近任意非线性函数。选用了应用最广泛的BP神经网络模型来预测黄金期货的价格。对采集到的影响黄金期货价格的因素数据进行了归一化处理后建立BP神经网络并进行了模拟训练,然后用训练好的网络进行检验,并比较了输出结果和真实值,发现用BP神经网络模型能够将误差控制在一个较小的范围内。经过实证研究可以发现BP神经网络用于价格预测可达到较好的效果。

Gold futures market is a very complex nonlinear system, however, the neural network has a strong capability of approximating any nonlinear system, theoretically, the neural network can infinitely approach the nonlinear function. So this arti- cle uses BP neural network to train and simulate the gold future price, then we use the neural network that has been trained well to obtain the forecast results. Finally, we compare the forecast results with the real results, then we find that the error is small when we use the BP neural network, and then gold future price of the next day is predicted using the trained neural net- work. After the analysis, we find that we can achieve good effect using the BP neural network to forecast in economy.