Application of Data-driven Method for Automatic Machine Learning in Economic Research

At present, the role of machine learning in data analysis is becoming increasingly important, and the digital economy has become the major economic form in the world, as well as the core driving force for China's economic development. Machine learning plays an increasingly significant role in economic research based on big data. To reduce the difficulty of using machine learning and improve the efficiency of machine learning, this paper systematically studies the application of automated machine learning (Au-toML) in economic research, focusing on the principles and characteristics of data-driven automated machine learning. Through the experimental comparison of specific automated machine learning methods on the classification of data sets, the optimal applicable method is found. Data-driven automated machine learning can be effectively applied in economic data mining, economic indicator analysis, and policy evaluation.