基于BP神经网络的图书馆读者流量预测  

Prediction of Library Reader Flow Based on BP Neural Network

作　　者：黄珍珍[1] 毛亚楠[2] 高萌[2] 

机构地区：[1]中国矿业大学图文信息中心,江苏徐州221116 [2]中国矿业大学计算机学院,江苏徐州221116

出　　处：《科教导刊》2012年第25期181-182,共2页The Guide Of Science & Education

摘　　要：本文主要基于BP神经网络对图书馆读者流量进行预测,建立BP神经网络模型,从输入层输入数据,逐层传递,将输出层输出的预测值与实际值比较得出相对误差,并以此为根据不断调整网络的权值和阈值,以获得最小误差。利用Matlab仿真模型对数据及预测结果误差进行分析,结果表明BP神经网络对图书馆读者流量进行预测具有较好的效果。This paper predicts library reader flow mainly based on BP neural network.In this paper,BP nerve network model is established.The data is input from the input layer and transferred in transfer layer.Then we can get relative error by comparing the prediction data from output layer to the actual value,and constantly adjust the weights of the networks and the threshold value to obtain the minimum error.Matlab is used to analyze the data and prediction results.The simulation results show that BP neural network is a effective method to predict library reader flow.

关 键 词：BP神经网络 读者流量预测 MATLAB仿真 

分 类 号：G250.72[文化科学—图书馆学]