【摘要】 BP神经网络方法在知识转移评价研究领域的运用普遍存在样本数量偏少、样本质量和代表性良莠不齐的问题。针对这一问题,本文在自建评价指标体系的基础上采用BP神经网络方法构建综合评价模型,通过大量发放企业问卷进行样本数据采集,用于神经网络的训练、检验及仿真。结果表明,BP神经网络评价模型更接近实际评价过程,能有效规避了常规综合评价方法中人为因素干扰及权重确定的主观性等弊端,从而具有较高的准确性和实用性。更多还原

【Abstract】 There are some deficiencies in evaluation study of knowledge transfer performance based on BP neural network:sample size too few and often associated with poor quality. Aiming at this problem, a comprehensive evaluation model wasdesigned with BP neural network method in accordance with self-established index system, and a large number of enter-prise questionnaires were distributed to collect sample data for neural network training, testing and simulation. The simula-tion result shows that the model has good generalization ability with high prediction and evaluation level, and also has ahigh degree of accuracy and operability.更多还原