基于改进BP神经网络的网络流量预测  

Network Traffic Prediction Based on Improved BP Neural Network

作　　者：吉珊珊[1] 柯钢[1] JI Shanshan;KE Gang(Department of Computer Engineering,Dongguan Polytechnic,Dongguan 523808)

机构地区：[1]东莞职业技术学院计算机工程系,东莞523808

出　　处：《计算机与数字工程》2020年第7期1682-1686,共5页Computer & Digital Engineering

基　　金：东莞职业技术学院示范建设专项资金项目(编号:政201819);2017广东省教育厅青年创新人才类项目(编号:2017GkQNCX116)资助。

摘　　要：为了提高网络流量的预测精度,克服BP神经网络预测过程中存在收敛速度慢、易陷入局部极小值的缺点,提出改进BP神经网络的网络流量预测模型。该模型引入动量因子和自适应学习速率来改进BP神经网络。仿真结果表明,改进BP神经网络预测的结果误差更小,精确度更高。In order to improve the prediction accuracy of network traffic and overcome the shortcomings of slow convergence speed and easy to fall into local minimum in the process of BP neural network prediction,an improved BP neural network traffic prediction model is proposed.This model introduces momentum factor and adaptive learning rate to improve BP neural network.The simulation results show that the improved BP neural network has smaller prediction error and higher accuracy.

关 键 词：BP神经网络 动量因子 自适应学习速率 网络流量 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]