AutoML: A Tertiary Study of Phases, Methods, Tools, and Frameworks

bstract. Automated Machine Learning (AutoML) plays a pivotal rolein making machine learning more accessible by automating key stepsin the model development process. Over the past decade, an increasingnumber of literature reviews (LRs) have examined specific components ofAutoML, including data preparation, feature engineering, model genera-tion, neural architecture search, hyperparameter optimization, and evalu-ation. However, a unified synthesis that consolidates findings across thesereviews is missing. This paper presents a tertiary review of 32 LRs toprovide a comprehensive and up-to-date overview of the AutoML land-scape. We systematically analyze the core AutoML phases, categorizeAutoML methods used across different stages of the machine learning(ML) pipeline, and compile a set of AutoML frameworks and tools. Thesynthesis offers a panoramic view of the techniques and tools supportingautomation across ML phases. Our findings aim to serve as a referencefor researchers and practitioners seeking to understand the current stateof AutoML, the extent of automation achieved across pipeline stages,and the tools and platforms that support these capabilities.Keywords: Automated Machine Learning (AutoML)· Tertiary Study ·Literature Review (LR) · AutoML Tools and Frameworks · Automationin Machine Learning (ML).