Deep-learning techniques have penetrated all aspects of our lives and brought us great convenience. However, the process of building a high-quality deep-learning system for a specific task is time-consuming, requires extensive resources and relies on human expertise, hindering the further development of deep learning applications in both industry and academia. To alleviate this problem, a growing number of research projects focus on automated machine learning (AutoML). In this paper, we provide a comprehensive and up-to-date study on the state-of-the-art (SOTA) in AutoML. First, we introduce the AutoML techniques in detail, in relation to the machine-learning pipeline. We then summarize existing research on neural architecture search (NAS), as this is one of the most popular topics in the field of AutoML. We also compare the performance of models generated by NAS algorithms with that of human-designed models. Finally, we present several open problems for future research.
View on arXiv