基于BP神经网络的图书馆用户参与度预测模型研究

Study on the BP Neural Network Model of Library User Participation Forecasting

本文依据不同服务模式下图书馆用户参与度的主要评价指标及影响因素,基于BP神经网络算法构建了图书馆用户参与度的预测模型。通过对某市公共图书馆用户参与度的预测与一元非线性回归预测模型对比验证了该模型具有更高的预测精度。分析认为:在全面把握影响用户参与外部条件与内部因素的基础上进行预测模型参数优化,可进一步提高预测结果的可信度。

Main evaluation indicators and influencing factors are discussed about library user participation under different services modes,and library user participation prediction and evaluation model is established by BP neural network algorithm.By predicting the user participation of certain city public library,it is proved that the prediction model has higher prediction accuracy compared with simple nonlinear regression model.Grasping the external conditions and internal factors of library participation comprehensively,and then optimizing the parameter of prediction model,which can further improve the reliability of predictions.