GAs and CA Basins of Attraction

From: Andrew Wuensche (100020.2727@CompuServe.COM)
Date: Mon Aug 02 1993 - 10:37:11 UTC


Further to recent mail on ..
Genetic Algorithms and Basins of Attraction of Cellular Automata.

The question may be, can basins of attraction (B-of-A), or the B-of-A
field, evolve towards some desired configuration by using a GA on a
population of CA rule tables. This may be possible - refer to "the Atlas"
pages 56-60, where mutation and cross-over of rules are discussed, and
appendix 3 (pages 225-233) which gives examples of CA rule mutants. 

The CA B-of-A/field  will not, however, be able to escape from the
general restrictions (due to various symmetries) described on pages 16-18
of "the Atlas".

Evolution towards some arbitrary "desired" B-of-A/field configuration
requires the relaxation of CA architecture, as described in "the Ghost in
the Machine", ie random Boolean networks (disordered CA). GA could be
applied to a population of random network parameters to evolve improved
parameters prior to learning (see page 26 of "the Ghost in the Machine").

The network is a dynamical system that categorises its state-space far
from equilibrium, as well as at the attractor. The global view of the
catagorisation is providerd by the B-of-A/field. This approach may
provide insights into the cognitive properties of networks (or networks
of networks).

Note that contrary to Ian Davidson's remark, these networks are strictly
synchronous, though asynchronous updating may be a possibility (see "the
Ghost in the Machine" page 19).


Andy Wuensche

contact address:                    Santa Fe Institute
48 Esmond Road, London W4 1JQ       and The University of Sussex (COGS)
tel 081 995 8893  fax 081 742 2178  wuensch@santafe.edu
100020.2727@compuserve.com          andywu@cogs.susx.ac.uk


THE GLOBAL DYNAMICS OF CELLULAR AUTOMATA
An Atlas of Basin of Attraction Fields of 
One-Dimensional Cellular Automata 
-----------------------------------------
Andrew Wuensche
Mike Lesser
Foreword by Chris Langton
Diskette included for PC-compatible computers.
Santa Fe Institute Studies in the Sciences of Complexity,
Reference Vol 1, Addison-Wesley, IBSN 0-201-55740-1.  


THE GHOST IN THE MACHINE
Basins of Attraction of Random Boolean Networks
-----------------------------------------------
Andrew Wuensche
Cognitive Science Research Paper 281, University of Sussex, 1993,
to be published in Artificial Life III, Santa Fe Institute Studies in the
Sciences of Complexity, Addison-Wesley. 
 


This archive was generated by hypermail 2.1.7 : Tue Oct 14 2003 - 21:44:19 UTC