摘要
提出了一种用户行为模式评估模型,此模型建立反向传播(BP)神经网络系统,对用户行为历史进行学习、训练、校验,最终得到稳定的神经网络形态。经过多重匹配和多次逼近后,成熟的神经网络系统模型可以实现对用户在时域与地域细分下的数据业务密度进行评估,在实际网络系统用户使用场景中,对用户分类和运营商策略给予基础理论参考和方法指引。
This paper presents an evaluation model of learning, training, and checking user behavior history, we approximation, the mature neural network system model cal subdivision, in the actual network system, it provides operator strategies. user behavior, and establish the BP neuralnetwork system, by get the stable form of the neural network. After multiple matching and can achieve the evaluation of data traffic density in time and geographi basic theory reference and guidance strategy for user classification and operator strategies.
出处
《中国新通信》
2014年第21期35-37,共3页
China New Telecommunications