BP神经网络优化网络科技文献共享评价体系

以网络科技文献共享的评价为对象,构建初选指标结构,对指标体系优化的主流评价方法进行研究及综述,在分析了各个评价方法的优劣的基础上确定运用BP神经网络的方法对初选的指标体系进行优化。然后基于BP神经网络对指标体系进行优化的方法及过程进行了论述。最后利用BP神经网络方法求取各项指标的权重,分析各个指标在整个指标体系中的重要程度,修改初选指标体系,进而达到优化的目的。

BP Neural Network Optimization Evaluation System of Network Sharing Scientific Literature

Making the evaluation of shared network science and technology literature as research object, the prima-ry index structure was built, the mainstream optimization evaluation methods of index system were studied and summarized,and the index system selected originally was optimized by using the BP neural network method based on the analysis of the advantages and disadvantages of various evaluation methods. The method and process of opti-mization on index system were also discussed based on BP neural network. Finally,the weight of each index was calculated by BP neural network method, the importance of degree of each index in the index system was analyzed and the primary index system was modified to achieve the goal of optimization.