Workshop on Entropy, Information and Complexity
Workshop on
Entropy, Information and Complexity
Organizers:
Stu Card, Critical Technologies Inc. & Syracuse University, US
Yossi Borenstein, University of Hertfordshire, UK
Workshop Description
Among the basic concepts and state variables of thermodynamics, ENTROPY plays a central role, explained by Boltzmann et al as statistical mechanics. Entropy and its complement [mutual] INFORMATION play key roles in the theory of communication by Shannon et al. COMPLEXITY and algorithmic information play the corresponding roles in the complexity theory of Kolmogorov et al.
- Boltzmann, Lotka, Odum and others have observed that the fundamental object of contention in the evolutionary struggle is available energy (free energy after discounting entropy). Ignorance of the thermodynamic laws that govern evolution does not enable evolutionary computation researchers to escape them, but does preclude our exploiting them.
- The vital essence of evolutionary learning consists of information flows between the natural or artificial environment and the entities differentially proliferating therein. Information theory enables objective measurement of these flows, of epistasis, and of gain or loss of information in individuals, ensembles and populations due to evolutionary steps, enabling design of EAs that efficiently 'distill' desired information from available data.
- Minimum Description Length (MDL) solutions satisfy Occam's Razor and are more likely to generalize well than are less parsimonious candidates. Normalized Compression Distance (NCD) and other computable estimates of non-computable complexity not only are theoretically justifiable but also have been shown to perform well in various real world applications such as classification trees.
All 3 of these theories are directly applicable to both natural biological evolution and evolutionary computation, but have generally been applied by neither theoreticians nor practitioners in the latter field. In recent years, a few papers have appeared, scattered across different venues. This workshop will bring together researchers with a common interest in applying these concepts systematically to evolutionary computation and learning.
Important Dates:
Submission: March 25, 2010
Notification: April 1, 2010
Camera ready: April 13, 2010
Submission:
This Call For Participation invites submissions of 8 page papers to be published in the GECCO proceedings and the ACM digital library (in ACM format, see http://www.sigevo.org/gecco-2010/papers.html) and extended abstracts of presentations on research in progress or concepts likely to be of interest to workshop participants. Each submitted paper will be independently reviewed by at least two reviewers. Upload your submission to this site or e-mail it to Stu Card.
