EMAA - Workshop on Empirical Methods for the Analysis of Algorithms

September 9, 2006
Reykjavik, Iceland

Held in conjunction with the
International Conference on Parallel Problem Solving From Nature (PPSN IX)

Proceedings - Workshop Description - Submission - Dates - Organising Committee - Program Committee


Luís Paquete, Marco Chiarandini, Dario Basso (Eds.): Workshop on Empirical Methods for the Analysis of Algorithms, EMAA 2006, Proceedings. Reykjavik, Iceland, September 9, 2006.
Content (inclusive of the slides of the four invited talks). BibTeX.

Workshop Description

It is currently acknowledged that randomised decisions in search algorithms, such as evolutionary algorithms and other metaheuristics, improve the capability of solving decision and optimization problems. The assessment of these algorithms through analytical studies is not trivial and often yields results of scarce importance for practical applications. Researchers tend therefore to report tables with results of empirical tests carried out on different problem instances and to comment on those results.

However, empirical analyses of algorithms require standard methods to guide in the design of the experiments and to determine whether the data collected are enough to draw any conclusions. The use of these empirical methods is not yet widely spread in the field and it is not yet fully understood which methods are appropriate for different scenarios of analysis, such as comparison, prediction and generalisation of algorithm performances.

This workshop aims at intensifying the discussion on empirical methods for the analysis of search algorithms, ultimately, contributing to the definition of a standard practice in experimental algorithmics.

Some of the topics are the following:

  • application of Experimental Design techniques, such as full or fractional factorial designs, response-surface methods, etc.;
  • statistical inference from computational results through parametric or non-parametric tests;
  • exploratory data analysis;
  • generalisation of algorithm complexity from empirical data;
  • analysis of run-time/solution-quality distributions;
  • definition of instance classes and benchmark problems and analysis of their representativeness;
  • tuning and sensitivity analysis of algorithm parameters;
  • impact of different performance measures on the characterisation of algorithm performance;
  • analysis of algorithm performance on stochastic and multiobjective problems;
  • influence of confounding factors such as hardware, compilers, data structures and implementation.

The focus will be on the methodologies, not on the performance of the algorithms themselves.


Authors should prepare their papers in the PPSN format by following the instructions of LNCS series of Springer. The length of the paper should not exceed 8 pages. Extended abstracts of post-deadline important results should be no longer than two pages, including figures and references. Accepted works will be presented at the Workshop.

Submissions should be made electronically, by sending papers or post-deadline abstracts in PDF format to: Luís Paquete lpaquete@ualg.pt or to Marco Chiarandini marco@imada.sdu.dk.

Important Dates

  • Deadline for paper submissions: 5th June, 2006
  • Notifications to authors: 25th June, 2006
  • Revised Manuscripts: 30 July, 2006
  • Post-deadline abstract submissions: 15th August, 2006
  • Post-deadline notifications: 31st August, 2006
  • EMAA workshop: 9th September, 2006

Organising Committee

Luís Paquete
Faculty of Economics and Centre for Intelligent Systems, University of Algarve

Marco Chiarandini
Department of Mathematics and Computer Science, University of Southern Denmark

Dario Basso
Department of Statistics, University of Padova

Program Committee

  • Thomas Bartz-Beielstein, Department of Computer Science, University of Dortmund
  • Mauro Birattari, IRIDIA, Université Libre de Bruxelles
  • Carlos M. Fonseca, Centre for Intelligent Systems, Faculty of Science and Technology, University of Algarve
  • Viviane Grunert da Fonseca, Instituto Superior D. Afonso III and Centre for Intelligent Systems, University of Algarve
  • Jürg Hüsler, Institute of Mathematical Statistics and Actuarial Science, University of Bern
  • Catherine C. McGeoch, Department of Mathematics and Computer Science, Amherst College
  • Fortunato Pesarin, Department of Statistics, University of Padova
  • Rubén Ruiz, Department of Applied Statistics and Operations Research and Quality, Valencia University of Technology
  • Luigi Salmaso, Faculty of Engineering, University of Padova
  • Thomas Stützle, IRIDIA, Université Libre de Bruxelles

Last modified: Wed Sep 27 09:00:03 CEST 2006