Scaling Behaviours of Landscapes, Parameters and Algorithms
Held in conjunction with the
Genetic and Evolutionary Computation Conference (GECCO 2011)
Organisers
Ender Özcan1, Andrew Parkes1 and Jon Rowe2
1. [>>] Automated Scheduling, Optimisation and Planning (ASAP) Research Group
[>>] School of Computer Science
[>>] University of Nottingham, UK
{exo, ajp} AT cs.nott.ac.uk
2. [>>] School of Computer Science
[>>] University of Birmingham, UK
J.E.Rowe AT cs.bham.ac.uk
Description | Workshop Format and Invited Speakers | Call for Abstracts | Important Dates
Description
All too often heuristics and meta-heuristics require significant parameter tuning to work most effectively. Often this tuning is performed without any a priori knowledge as to how good values of parameters might depend on features of the problem. This lack of knowledge can lead to lot of computational effort and also has the danger of being limited to only problem instances that are similar to those that have been seen before. The aim of the workshop is to develop methods to give deeper insight into problem classes, and how to obtain and exploit structural information. In particular, we often would like to be able to tune parameters using small instances (for speed) but then adjust so as to be able to run on large instances. This will require some theory of how to extrapolate tuning outside of the size or features of the training set. An analogy is the difference between non-parametric and parametric statistics; the former does not assume any underlying probability distribution and the latter can (for example) assume a Gaussian. Naturally, the latter might give stronger results and with smaller sized samples. Hence, to distinguish this from standard parameter tuning, we might call this "Parametric Parameter Tuning". Of course, this is a challenging problem; but we hope to be able to discuss any existing work and how the community might meet the challenge.
Related to this is the common and natural belief that the semantic properties of the landscapes will be reflected in the performances of algorithms. A subsequent underlying assumption, or hypothesis, if the landscape has a particular functional dependence on features of the instance, then such functional dependencies are also likely to play a key role in understanding the behaviour of heuristic algorithms, and so merit investigation. We are particularly interested the area of phase transitions; when particular semantic properties display phases of 'almost always true' and 'almost never true'. Statistical methods can then reveal some appropriate parameters to describe the locations of such phases, and we expect that this will also influence the understanding, design and tuning of algorithms. This is exemplified by the work in the artificial intelligence and statistical physics communities on propositional satisfiability and graph colouring, and that has led to deeper understanding of algorithms, and development of new ones. One of the goals of the workshop is to look into phase transition theory with a view to potential applications to traditional GECCO problems.
The target participants are those that:
· Work on the theory of search algorithms, but are seeking ways for the theory to have a practical impact
· Work on direct applications, but are frustrated with the trial-and-error approaches that often are
often used, and would like to bring 'theoretically-inspired methods' into their work.
We also aim to bring together researchers and practitioners from related fields such as Operational Research (OR), Artificial Intelligence and Computer Science, providing a medium for sharing and inspiring of techniques (even if application domains are different) and developing common understandings.
Workshop Format and Invited Speakers
We have a two hour workshop.
It is our intention to have a rich and lively workshop made up of a variety of views and perspectives.
ROOM: Elgin
Tentative programme:
|
Time |
Authors |
Title |
|
08:30 – 08:45 |
Organisers |
Introduction |
|
08: 45– 09:00 |
Valentino Santucci, Alfredo Milani |
Covariance-based Parameters Adaptation in Differential Evolution |
|
09:00 – 09:15 |
Carolina Salto, Enrique Alba, Francisco Luna
|
Using landscape measures for the online tuning of heterogeneous distributed GAs |
|
09:15 – 09:30 |
P. Caamańo, J.A. Becerra, F. Bellas, R.J. Duro |
Are evolutionary algorithm competitions characterizing landscapes appropriately? |
|
09:30 – 09:45 |
Tianjun Liao, Marco A. Montes de Oca, Thomas Stützle |
Tuning Parameters across Mixed Dimensional Instances: A Performance Scalability Study of Sep-G-CMA-ES |
|
09:45 – 10:00 |
David Corne, Alan Reynolds |
Evaluating optimization algorithms: bounds on the performance of optimizers on unseen problems |
|
10:00 – 10:20 |
Participants |
Discussion |
[Top]
Call for Abstracts
(Tentative details) We invite submissions as extended abstracts of 3-4 pages without exceeding 1000 words in ACM format. Please see the GECCO 2010 information for authors for further details. However, unlike GECCO, papers do not have to be submitted in anonymous format. The abstracts should be submitted in PDF format directly to the following e-mail address: ajp ATT cs . nott . ac . uk, and please mention “GECCO HU submission” in the subject line.
All accepted papers will be presented at the workshop and appear in the GECCO workshop volume. Proceedings of the workshop will be published on CD-ROM, and distributed at the conference.
[Top]
Important Dates
· Deadline for abstract submission: 7 April, 2011
· Notification of Acceptance: 14 April, 2011
· Camera-ready deadline: 26 April, 2011
· Registration deadline: 2 May, 2011
[Top]
Ender Özcan and Andrew J. Parkes are members of [the LANCS initiative].