Evolving Fuzzy Systems --- Fundamentals, Reliability, Interpretability, Useability and Applications

This chapter provides a round picture of the development and advances in the field of evolving fuzzy systems (EFS) made during the last decade since their first time appearance in 2002. Their basic difference, opposed to conventional fuzzy systems (discussed in other chapters in this book), is that they can be learned from data on-the-fly during (fast) on-line processes in an incremental and mostly single-pass manner. Therefore, they stand for a very emerging topic in the field of soft computing for addressing modeling problems in the quickly increasing complexity of real-world applications, more and more implying a shift from batch off-line model design phases (as conducted since the 80ties) to permanent on-line (active) model teaching and adaptation. The focus will be placed on the definition of various model architectures used in the context of EFS, providing an overview about the basic learning concepts and listing the most prominent EFS approaches (fundamentals), discussing advanced aspects towards an improved stability, reliability and useability (usually must-to-haves to guarantee robustness and userfriendliness) as well as an educated interpretability (usually nice-to-haves to offer insights into the systems' nature). It will be concluded with a list of real-world applications and challenges where various EFS approaches have been successfully applied with satisfactory accuracy, robustness and speed.