基于深度学习的通信网络数据关键特征挖掘

Key Feature Mining of Communication Network Data Based on Deep Learning

深度学习是挖掘数据关键特征的重要技术手段,为准确分析通信网络数据特征,并保障质量,提出基于深度学习的通信网络数据关键特征挖掘方法。选取接入率、可用性以及覆盖率等七个指标作为通信网络质量核心性能指标,将卷积神经网络与径向基神经网络相结合,构建深度学习网络结构,将该性能指标作为标签参数,将所得到的标签参数的聚类与求和结果作为深度网络的标签数据,通过前向传播将标签数据输入卷积神经网络的输入层内,经过不同隐层的变换与映射至输出层位置,并采用量子粒子群算法求解深度学习网络最优参数,输出通信网络数据关键特征挖掘结果。经实验结果表明,所提方法的通信网络数据关键特征挖掘率在95%以上,能够准确预测未来短时间段内的通信网络质量。

In order to accurately analyze the data characteristics of communication network and ensure the quality of communication network,a key feature mining method of communication network data based on deep learning is proposed.Seven indicators such as access rate,availability and coverage are selected as the core performance indicators of communication network quality.The convolutional neural network and radial basis function neural network are combined to construct the deep learning network structure.The core performance indicators of communication network quality are taken as the label parameters,and the clustering and summation results of the above label parameters are taken as the label data of the deep network.The label data is input into the input layer of convolutional neural network through forward propagation,transformed and mapped to the position of output layer through different hidden layers,and the quantum particle swarm optimization algorithm is used to solve the optimal parameters of deep learning network,and the key feature mining results of communication network data are output.The experimental results show that the key feature mining rate of communication network data is more than 95%and can accurately predict the communication network quality in a short period of time in the future.