Data-driven approach with artificial neural network for Global Innovation Index re-evaluation

The Global Innovation Index (GII) is employed to provide rankings of innovation capabilities for all countries and economies around the world. To consider the internal relationship between factors and monitor the change of innovation capability level in real-time, a data-driven approach based on the random forest and artificial neural network is proposed in this study. It can estimate the Global Innovation Index for every country or economy quickly and effectively under the condition of mastering a few indicators. Taking the 2019 and 2020 GII data for a case study, the results show that the approach can effectively calculate the approximate scores of the countries and economies using only 14 indicators selected from the GII reports. The approach provides an opportunity for countries and economies all around the world to monitor their development using indicators from these specific few fields. And further to improve national innovative capacity with the feedback knowledge mined from the data.