基于神经网络模型的地方智库竞争力评估——以江苏淮安地方智库为例

Competitiveness Evaluation of Local Think Tank Based on Neural Network Model——A Case of the Local Think Tank in Huai'an

基于战略管理理论、竞争力分析理论和计算实验逻辑框架,以江苏淮安若干地方智库为研究对象,运用神经网络模型(包括三层BP神经网络模型和Elman神经网络模型)、遗传算法对地方智库进行竞争力评估.研究结果表明:运用遗传算法约简得到的少数评估指标可以有效地评估智库竞争力,以上两种神经网络模型均可适用于地方智库的评估工作,Elman模型相对略优.

According to strategic management theory, competitiveness analysis theory and computational experiments logical framework, it carried through competitiveness evaluation on local think tanks from Huai＇an city of Jiangsu province, in use of neural network（ including three layer BP neural network model, Elman neural network model ） and genetic algorithm. The research results show that a few evaluation indices reduced by genetic algorithm can be used to evaluate think tank competitiveness effectively, and the above two kinds of neural network models are both useful for evaluation on local think tank, but Elman neural network model is something better.