|
Home
Call for Papers
Workshops
Program
Program Committee
Keynote Speakers
Submissions
Registration
Lodging
Travel
Sponsors
Artwork
by Dr. Cliff Pickover used by permission
|
|
Alife 9 tutorial
Part I: Introduction to Random Boolean Networks
Carlos Gershenson
Centrum Leo Apostel, Vrije Universiteit Brussel
Krijgskundestraat 33, Brussels, 1160, Belgium
cgershen@vub.ac.be
http://homepages.vub.ac.be/~cgershen
and
Part II: Studying discrete dynamical
networks with DDLab
Andrew Wuensche
UWE and Sussex University
andyw @ cybermesa.com
PART I: This will be a short tutorial to the artificial life community to promote
interest in
the study of Random Boolean Networks (RBNs), a class of discrete dynamical
network. These can be very interesting
tools, since one does
not have to assume any functionality or particular connectivity of the networks
in order to study their
generic properties. Like this, RBNs have been used for exploring the
configurations where life could
emerge. The fact that RBNs are a generalization of cellular automata makes their
research a very
important topic for artificial life.
PART II: An introduction to Discrete Dynamics Lab (DDLab),
an interactive graphical software tool for studying cellular automata, random
Boolean networks, or any intermediate hybrid including scale-free networks, or
networks of subnetworks. There will be concrete illustrations of many of the
themes introduced in the first part of this tutorial.
Part I Outline
Introduction to RBNs, Carlos Gershenson
The intention of Part I of the tutorial is to present the state of the art in RBNs,
spanning over several lines of research carried out by different groups. This
will last two hours, with
the following outline:
Introduction
Classical model (Kauffman, 1969; 1993), genetic regulatory networks
-Transitions between order and chaos (Derrida and Pomeau,
1986)
-Explorations (Wuensche, 1997; Aldana et al., 2003)
- Multi-valued networks (Solé. Luque, and Kauffman, 2000)
- Topologies (Oosawa and Savageau, 2002; Aldana, 2003)
- RBN control (Ballesteros and Luque, 2002)
ALife criticisms:
" Asynchronous RBNs (Harvey and Bossomaier, 1997)
" Rhythmic Asynchronous RBNs (Di Paolo, 2001; Rholfshagen and Di Paolo, 2004)
" Deterministic Asynchronous RBNs (Gershenson, 2002)
" Mixed-context RBNs (Gershenson, Broekaert and Aerts, 2003)
Different updating schemes
" classification (Gershenson, 2002)
" statistics
- attractors
- percentage of states in attractors
" complexity reduction (Gershenson, 2004)
" phase transitions (order-chaos) (Mesot and Teuscher, 2003; Gershenson,
submitted)
Future lines of research
Conclusions
Tools
Part II Outline
DDLab, Andrew Wuensche
Discrete Dynamics Lab is interactive graphical software for studying cellular
automata, random Boolean networks, or any intermediate hybrid including
scale-free networks, or networks of subnetworks. Nodes change their value
according to inputs from other nodes in the network according to Boolean or
multi-value logic (up to 8 values/colors). Networks can be easily created,
visualized and amended. DDLab then studies the dynamics ON the network:
space-time patterns (in 1d, 2d and 3d) and basins of attraction.
I will show how to operate DDLab in a live, interactive tutorial, and
describe various ideas, methods and results, including: categorization (memory)
in basins of attraction and subtrees, the consequences of
mutations/perturbations, and for CA in particular: order-complexity-chaos
measures and parameters, the automatic classification of rule-space by
input-entropy, encryption, the emergence of gliders, self-reproduction by
collisions, and filtering the gliders' background. DDLab runs in Linux, Unix,
Mac, Irix and DOS. For more information see
http://www.ddlab.com .
|
