基于人工蜂群改进的BP神经网络移动用户行为分析及预测方法

Improved BP Neural Network Algorithm Based on Artificial Bee Colony of Mobile User Behavior Analysis and Forecasting

如何根据不同的用户行为,来为移动用户提供精准的个性化服务是目前移动应用服务开发技术发展的主流。为解决BP神经网络建模算法收敛速度慢及预测不准确问题,提出基于人工蜂群算法改进的BP神经网络算法。为测试改进后算法的准确性,采用Matlab编程进行试验仿真,通过黑盒子测试方法输出预测的用户行为和实际的用户行为。在18次预测中只有2次预测失败,预测成功率达80%以上。为了验证改进的BP神经网络算法的效率,采用初始总群数为1000,进行了收敛性测试。试验结果表明:基于人工蜂群算法改进的BP神经网络算法可以有效的提高移动用户行为分析的效率和准确性,对在使用移动用户行为分析模型构建过程中,准确定位用户上网需求,提升企业在营销中的竞争力具有非常重要的意义。

It is the mainstream of development technology for current mobile application service to provide accurate and personalized service for mobile users based on different users＇ behavior. This paper proposes an algorithm to improve BP neural network based on artificial bee colony in order to solve BP neural network algorithm slow convergence and inaccurate prediction. To test the accuracy of the improved algorithm, this paper uses Matlab programming experiment simulation. This paper outputs prediction and the actual user behavior through the black box testing method. Only twice were failed in 18 times forecast and the success rate was more than 80%.In order to validate the efficiency of the improved BP neural network algorithm, this paper tests the convergence by initializing total group number 1000. Results showed that the improved BP neural network algorithm based on artificial colony algorithm could effectively improve the efficiency and accuracy of the mobile user behavior analysis. It is very important to locate user＇s demand to Internet accurately and promote the power of competition in the marketing of enterprises during the process of the construction of analysis model based on the mobile users＇ behavior.