Analysis and Evaluation of Innovation Performance of Micro, Small and Medium-sized Technology Enterprises Based on ANN-RBF

The survival of micro, small and medium-sized technology enterprises is highly dependent on innovation, while they also have strong innovation ability and development potential. Based on summarizing the existing research methods, this paper adopts the general, objective, and available indicators to quantify innovation and innovation performance. The RBF (Radial Basis Function) neural network is established with an average overall relative error of 1.271 through multiple pieces of training to analyze the selected sample data. The result shows that the proportion of researchers whose importance of normalization is 100.0 % and the cooperation orientation with normalization is 88.2 %, which significantly impacts innovation performance. The research of this paper provides scientific theoretical guidance for micro, small and medium-sized technology enterprises to rationally allocate innovation input factors and effectively improve their innovation performance.