Recent Developments of Automated Machine Learning and Search Techniques

Abstract

The recent successes of artificial intelligence, in particular machine learning, for solving real-world problems have motivated the advances towards automated design of algorithms and systems with less human involvement. In machine learning and meta-heuristic search algorithms, different lines of relevant research are now emerging, with findings feeding into each other. This book presents a selection of some recent advances across automated machine learning (AutoML) and automated algorithm design (AutoAD), where the effectiveness and efficiency of techniques and algorithms has been enhanced with the support of new taxonomies, models, theories, as well as frameworks and benchmarks. The emerging new lines of exciting research directions in AutoML and AutoAD present new challenges across multiple research communities in machine learning, evolutionary computation and optimisation research.