使用受限玻尔兹曼机和反向传播(BP)神经网络相结合的方法, 充分利用受限玻尔兹曼机自动提取数据的内在特征的能力, 以及 BP神经网络的学习能力, 对已建立的装备保障评价体系进行评估。利用 Matlab对方法的有效性进行验证, 通过公共数据集进行仿真。结果表明, 用该方法获得的评估值精确度较高, 具有良好的识别率。

The method combing the Restricted Boltzmann Machine(RBM) with the Back Propagation(BP) neural network is employed to assess the established equipment support evaluation system. This method makes full use of the ability of automatically extracting the inherent characteristics of the data for RBM, and the learning ability of BP neural network. Using Matlab to validate the effectiveness of the method, the simulation results show that the proposed method can realize high precision assessment of the value with a good recognition rate for the public data sets.