| Summary: | China’s high-tech parks have significant effects on driving national ecological innovation. Among them, ten world-class high-tech parks represent the highest level of development in China’s high-tech industry. Understanding the development characteristics of national world-class high-tech parks is of great significance for guiding the construction of other parks and achieving the high-quality development of parks. Based on the evaluation data of over 200 indicators of national high-tech parks from 2013 to 2017, this study used the XGBoost classic machine learning algorithm to select the characteristic indicators of national world-class high-tech parks and establish an evaluation indicator system, and it identified four primary indicators of the world-class high-tech parks, including innovation development, enterprise development, international development, and economic development. The indicators cover 30 important sub-indicators and highlight the importance of innovation resource input indicators, such as “use of technology activity funding from government departments”, “full-time equivalent of R&D personnel”, and “financial technology expenditure in high-tech parks”. Compared to the expert analysis, the application of the machine learning method in the evaluation of national high-tech parks improves the efficiency of selecting important indicators and makes the selection results more objective. The results of this research provide a reference value for guiding and promoting national high-tech parks to become world-class parks.
|
|---|