CEC 2013 Fiesta Americana Grand Coral Beach Hotel Grand Coral Ballroom Chichen Itza and Tulum Isla Contoy Isla Mujeres Ruinas del Rey Cozumel and Hel-Ha Kayaking and Windsurfing Tres rios and Actun Chen

Tutorial

An Introduction to Statistical Analysis of EC Experiments

Mark Wineberg
Associate Professor School of Computer Science
University of Guelph, Ontario, Canada

Abstract:

Although the situation has been improving over the past few years, many researchers in the EC field are still using an unfortunate lack of statistical rigor when performing and analyzing experiments. Experimental results that lack a proper statistical analysis should be considered as anecdotal at best, and may even be wholly inaccurate. This tutorial aims at providing the basic statistical framework that all experimenters in the EC field should follow.

The tutorial will cover introductory topics in Statistics such as the T test, confidence intervals, the Central Limit theorem (its power and limitations), non-parametric statistics, and will touch on multi-variate and multi-factorial analysis such as regression analysis, both linear and polynomial, as well as Analysis of Variance (ANOVA). The tutorial will be presented at an introductory level and is meant for any EC researchers who want to compare their newly designed EC system with established EC systems to see if there is an improvement, or who want to determine which EC system performs better on their chosen problem; i.e. nearly everyone. It is vital, if our community is to be taken seriously, for us to continue to educate ourselves (especially new Graduate students) on the statistical techniques that are considered de rigor for experiments performed in any field, such as ours, that is stochastic in nature.

Expected Enrolment:

A much simpler version of this tutorial was given by myself and Steffen Christensen at the CEC in Portland OR in 2004 and again in Edinburgh in 2005. The tutorial has been given yearly at GECCO from 2004 on, and will be given again this year in its current form (which now includes regression and ANOVA). The tutorial was well attended when at CEC in 2004-2005 drawing about 80-120 members. At GECCO, attendance fluctuates depending on the year. Some years it can have up to 200 members in the audience, but it has never been lower than 30.
 

Biography:

Prof. Wineberg has been actively researching the field of Evolutionary Computation (EC) and Genetic Algorithms (GA) since 1993 while he was still a graduate student at Carlton University, Ottawa, Canada. Over the years he has published on various topics including: the intersection of Genetic Algorithms and Genetic Programming, enhancing the GA for improved behavior in dynamic environments through specialized multiple populations, and exploring the concept of distances and diversity in GA populations. He is currently continuing his research in dynamic environments and studying the properties that allow one population to evolve more readily than another. Prof. Wineberg also teaches an undergraduate course at Guelph on computer simulation and modeling of discrete stochastic systems, which emphasizes proper statistical analysis, as well as a graduate course on experimental design and analysis for computer science.

 

The length of the tutorial:
two hours.

//

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer