Book 'Self Adaptive, Evolving Systems'

In today’s real-world applications, there is an increasing demand of integrating new information and knowledge on-demand into model building processes to account for changing system dynamics, new operating conditions, varying human behaviors or environmental influences. This increases the efficiency of models in industrial systems with respect to performance and process safety during on-line operation and production phases.
Evolving fuzzy systems (EFS) are a powerful tool to cope with this requirement, as they are able to automatically adapt parameters, expand their structure and extend their memory on-the-fly, allowing on-line/real-time modeling.

This book comprises several evolving fuzzy systems approaches which have emerged during the last decade and highlights the most important incremental learning methods used. The second part is dedicated to advanced concepts for increasing performance, robustness, process-safety and reliability, for enhancing user-friendliness and enlarging the field of applicability of EFS and for improving the interpretability of the evolved models. The latter highlights concepts and aspects regarding the question whether the evolved models are in fact understandable by human beings. An answer to this question could be a solid basis for an enhanced communication and interaction between man and machine in the near future.

The third part underlines the usefulness and necessity of evolving fuzzy systems in several online real-world application scenarios including system identification, on-line quality control, visual inspection, human-perception based texture analysis, chemometric modeling, provides an outline of potential future applications and raises open problems and new challenges for the next generation evolving systems, including human-inspired evolving machines and hybrid active learning (HAL).

The book includes basic principles, concepts, algorithms and theoretic results underlined by illustrations. It is dedicated to researchers from the field of fuzzy systems, machine learning, data mining and system identification as well as engineers and technicians who apply data-driven modeling techniques in real-world systems.

About the Author

Current Affiliaton: Post-Doc at the Department of Knowledge-Based Mathematical Systems, University of Linz
Expertise: fuzzy systems, incremental learning, evolving systems, machine learning, data mining, fault detection, quality control, texture analysis

Project Experience: 3 EU-Projects (AMPA, DynaVis, SynTeX) in the field of machine vision, data-driven modeling, quality control as key researcher of JKU

Research Activities: organizations of diverse special sessions and issues at various conferences and journals, published more than 50 papers in the field of evolving fuzzy systems, machine learning, quality control, visual inspection,
contributes as reviewer for various international journals

Grants:
- Best Paper Award 2006 at the Evolving Fuzzy Systems Workshop 2006 in Ambleside (Lake District)
- Royal Society Grant 2007 for Know-How Exchange with University of Lancaster
- Best Paper Award 2008 at the Genetic and Evolving Fuzzy Systems Workshop
2008 in Dortmund

Current Work:
- Key Researcher in the K-Project PAC (Process Analytic Chemistry)
- Initiatior of the bilateral FWF-DFG Project 'Interpretable and Reliable EFS'

Dr. Edwin Lughofer
Fuzzy Logic Laboratorium Linz-Hagenberg
Softwarepark 21
A-4232 Hagenberg
http://www.flll.jku.at/staff/edwin

This release was published on openPR.