A Survey of Empirical Research on Deep Learning in China

Domestic Research on deep learning is a multi-dimensional and in-depth development trend. Reviewing existing research in a timely manner is of great practical significance to promote the comprehensive development of deep learning theory and practice. This article adopts a research method that combines literature research and content analysis. This paper analyzes the domestic deep learning empirical research literature of the China National Knowledge Infrastructure (CNKI) from the past five years. The results showed three primary findings. (1) The research focuses on the design, implementation, and influencing factors of the deep learning process. (2) The theoretical reasoning of empirical research is mainly oriented to three deep learning meanings: learning method, learning process, and learning ability. (3) The investigation method, the experimental method, the construction method, and the case method are the common methods of empirical research. Self-report, observational evaluation, and automatic measurement are the main methods of deep learning measurement.