为了克服神经网络依赖初始化结果,泛化能力不强的缺点,提出一种基于受限玻尔兹曼机(RBM)的神经网络模型.利用无监督学习方法优化神经网络的初始权值和阈值,将RBM与神经网络融合起来,模型与时间序列神经网络做实验对比,结果表明,基于受限的玻尔兹曼机的神经网络模型优于神经网络预测模型,模型可以提高预测的精准度,具有一定的应用意义. In order to overcome the disadvantage of neural network relying on initializing results and poor generalization ability, a neural network model based on constrained Boltzmann machine (RBM) is proposed.Using unsupervised learning method to optimize the initial weights and thresholds of neural networks, The results show that the neural network model based on the limited Boltzmann machine is superior to the neural network prediction model, the model can improve the accuracy of prediction, and has the advantages of A certain significance of the application.