Abstract: Automated feature engineering is an efficient feature engineering solution that does not require human intervention, and is a popular research topic in machine learning research. In order to help the users employ automated feature engineering algorithms to create high-quality feature sets, this paper discussed the technics of current automated feature engineering frameworks from feature generation, feature selection, and system implementation, summarized, analyzed, and summarized the shortcomings of the existing methods. On the basis of existing analysis, we pointed out the future directions and challenges of automated feature engineering.