School of Management, Shenyang University of Technology, Shenyang, China.
Dean's Office, Liaoning Engineering Vocational College, Tieling, China.
PLoS One. 2022 Aug 25;17(8):e0271316. doi: 10.1371/journal.pone.0271316. eCollection 2022.
Policy synergy is necessary to promote technological innovation and sustainable industrial development. A radial basis function (RBF) neural network model with an automatic coding machine and fractional momentum was proposed for the prediction of technological innovation. Policy keywords for China's new energy vehicle policies issued over the years were quantified by the use of an Latent Dirichlet Allocation (LDA) model. The training of the neural network model was completed by using policy keywords, synergy was measured as the input layer, and the number of synchronous patent applications was measured as the output layer. The predictive efficacies of the traditional neural network model and the improved neural network model were compared again to verify the applicability and accuracy of the improved neural network. Finally, the influence of the degree of synergy on technological innovation was revealed by changing the intensity of policy measures. This study provides a basis for the relevant departments to formulate industrial policies and improve innovation performance by enterprises.
需要政策协同以促进技术创新和可持续产业发展。本文提出了一种具有自动编码机和分数动量的径向基函数（RBF）神经网络模型，用于预测技术创新。利用潜在狄利克雷分配（LDA）模型对中国历年新能源汽车政策中的政策关键词进行量化。通过使用政策关键词完成神经网络模型的训练，将协同度作为输入层，将同步专利申请数量作为输出层。再次比较传统神经网络模型和改进后的神经网络模型的预测效果，验证改进后的神经网络的适用性和准确性。最后，通过改变政策措施的强度来揭示协同度对技术创新的影响。本研究为相关部门制定产业政策和提高企业创新绩效提供了依据。
