Donghua University

In this tutorial, we will summarize the exciting evolution of evolutionary strategies – a branch of evolutionary computation which has introduced important ideas such as self-adaptation, emphasis on mutation operators, environmental selection, small population sizes, to mention only a few, into the field. Most of the basic algorithmic variants of evolutionary strategies will be presented in the tutorial, such that attendees will understand the concepts and will be able to implement the algorithms by themselves.

Some recent developments such as the introduction of niching methods to evolutionary strategies for discovering multiple optima within one run, and the development of the Mixed-Integer evolutionary strategy (MIES), will also be presented, with corresponding application examples such as the optimization of femtosecond laser pulses and of intravascular ultrasound image classification, as well as some application in drug design.

Some basic theory regarding convergence velocity and convergence reliability of evolutionary strategies will complete the picture, such that attendees will get a feeling for all aspects of evolutionary strategies.

No special initial knowledge about evolutionary computation is required to attend this tutorial.

Bio:

Thomas Bäck, PhD, is Professor for Natural Computing at the Leiden Institute for Advanced Computer Science (LIACS) at Leiden University, The Netherlands, and Chief Scientist of the Qe3 group at Netezza Corporation. Thomas received his PhD in Computer Science from Dortmund University, Germany, in 1994. Thomas Bäck has more than 150 publications on natural computing technologies, as well as a book on evolutionary algorithms, entitled Evolutionary Algorithms: Theory and Practice.

He is co-editor of the Handbook of Evolutionary Computation, and of the new Handbook of Natural Computing (to appear in 2010). He is editorial board member and associate editor of a number of journals on evolutionary and natural computation, and has served as program chair for all major conferences in evolutionary computation. His expertise lies in adaptive technologies for optimization and data-driven modeling, predictive analytics, and bioinformatics. He received the best dissertation award from the Gesellschaft für Informatik (GI) in 1995 and is an elected fellow of the International Society for Genetic and Evolutionary Computation for his contributions to the field.

Thomas has ample experience in applying evolutionary computation to business problems, working with Fortune 500 customers such as e.g. Audi, Air Liquide, BMW, Corning, Daimler, Henkel, Honda, Johnson & Johnson, P&G, Unilever, Volkswagen.

Abstract:

Grammatical Evolution is an automatic programming system that is a form of Genetic Programming that uses grammars to evolve structures.
These structures can be in any form that can be specified using a grammar, including computer languages, graphs and neural networks.
When evolving computer languages, multiple types can be handled in a completely transparent manner.

This tutorial gives a brief introduction to Backus Naur Form grammars and a background into the use of grammars with Genetic Programming, before describing the inner workings of Grammatical Evolution and some of the more commonly used extensions.

Bio:

Dr. Conor Ryan is a Senior Lecturer and University Fellow at the University of Limerick in Ireland, where he is also director of the Biocomputing and Developmental Systems (BDS) group. He is the inventor of Grammatical Evolution, and the BDS developed a tool, libGE, for evolving grammars that was recently voted the most commonly used Genetic Programming tool on the GP Mailing List.

He is active in many real world applications, such as semi-automated Mammography, Financial Prediction and Telecommunications. He is a director and co-founder of Evolvability, a high tech start up company that tests and optimizes Flash Memory and Solid State Device Disks.

Genetic Programming (GP) is a complex adaptive system with an immense number of degrees of freedom. Understanding how, why and when it works is difficult. Its behaviour is typically investigated in two ways: experimentally and theoretically.

Experimental studies require the experimenter to choose which problems, parameter settings and descriptors to use. Plotting the wrong data increases the confusion about GP's behaviour, rather than clarify it.

A more powerful alternative is to study GP theoretically. In this tutorial we will look at GP as a search process and explain its behaviour by considering the GP search space, in terms of its size, its limiting fitness distributions and also the halting probability. We will then use modern schema theory to characterise GP search. Finally, we will be in a position to explain the reasons for spurious phenomena (such as bloat) in GP and we will look at theoretically-sound ways of curing them.

Some prior knowledge of GP will be assumed.

Background reading:

* W.B. Langdon and R. Poli, Foundations of Genetic Programming, Springer, 2002.

* R. Poli, W.B. Langdon and N.F. McPhee, A Field Guide to Genetic Programming, Lulu.com, 2008 (freely available from the Internet).

Bio:

Riccardo Poli is a Professor in the School of Computer Science and Electronic Engineering of the University of Essex in the UK. He started his academic career as an electronic engineer doing a PhD in biomedical image analysis to later become an expert in the field of EC. He has published around 270 refereed papers and two books on the theory and applications of genetic programming, evolutionary algorithms, particle swarm optimisation, biomedical engineering, brain-computer interfaces, neural networks, image/signal processing, biology and psychology.

He is a Fellow of the International Society for Genetic and Evolutionary Computation (since 2003), a recipient of the EvoStar award for outstanding contributions to this field (2007), and an ACM SIGEVO executive board member (2007-2013). He was co-founder and co-chair of the European Conference on GP (1998-2000, 2003). He was general chair (2004), track chair (2002, 2007), business committee member (2005), and competition chair (2006) of ACM's Genetic and Evolutionary Computation Conference, co-chair of the Foundations of Genetic Algorithms Workshop (2002) and technical chair of the International Workshop on Ant Colony Optimisation and Swarm Intelligence (2006). He is an associate editor of Genetic Programming and Evolvable Machines and the International Journal of Computational Intelligence Research. He is an advisory board member of the Evolutionary Computation Journal and the Journal on Artificial Evolution and Applications. He is an editorial board member of Swarm Intelligence. He is a member of the EPSRC Peer Review College, an EU expert evaluator and a grant-proposal referee for Irish, Swiss and Italian funding bodies.