Cooperation Network and Enterprise Key Core Tech-Nology Breakthrough

Based on random forest model

Abstract

The impact of cooperative network on enterprise innovation performance has been paid attention by the academic community. The importance evaluation of cooperative network on enterprise key core technology breakthroughs is the focus  of this paper. This paper takes China's high-tech manufacturing industry from 2010 to 2020 as a research sample to build a cooperative network, uses a random forest-based machine learning model to measure the im-portance of cooperative network to the key core technology breakthroughs of enterprises, and further analyzes the relative importance of cooperative network features to fit the key core technology breakthroughs of enterprises. The findings are as follows: (1) Overall, the lack of cooperative network features will reduce the fitting performance of key core technology breakthroughs. (2) Rank the features in the cooperative network according to the degree of im-portance, and further use the partial dependence diagram to visualize the marginal impact of the three most important features, individual network size, number of intermediaries and proximity centrality, on the breakthrough of key core technologies of enterprises.