Information Technology Reference
In-Depth Information
create and move structures, as well as link up existing ones. Fractals (Mandelbrot
1983
; Fractal Foundation
2014
) are also important and cover natural systems and
chaos theory. There are many examples of fractals in nature. Using a relatively
simple
, the complex systems that actu-
ally exist can be created. As described in Wolfram (
1983
), automata and fractals
share the feature of self-similarity, where portions of the pattern, if magni
'
feedback with minor change mechanism
'
ed, are
indistinguishable from the whole. Tree and snow
ake shapes can be created using
fractals, for example. Fractals also show how well de
fl
ned these natural non-bio
processes are already. So automata would belong to the group called fractals and are
created using the same types of recursive feedback mechanism. The construction of
a concept tree would be a self-repeating process, but the created structures are not
self-similar. However, they would result from same sort of simplistic feedback
mechanism that these self-similar systems use.
Agent-Based modelling is another form of distributed and potentially intelligent
modelling. Scholarpedia
4
notes that Agent-Based Models (ABM) can be seen as the
natural extension of the Ising model (Ising
1925
) or Cellular Automata-like models.
It goes on to state that one important characteristic of ABMs, which distinguishes
them from Cellular Automata, is the potential asynchrony of the interactions among
agents and between agents and their environments. Also ABMs are not necessarily
grid-based nor do agents
'
'
the environment. An introduction to ABM could be
the paper Macal and North (
2006
). Agent-based models usually require the indi-
vidual components to exhibit autonomous or self-controlled behaviour and to be
able to make decision for themselves, sometimes pro-actively. While Cellular
Automata would be considered too in
tile
exible, agents would probably be considered
as too sophisticated. Although as noted in Macal and North (
2006
), some modellers
consider that any individual component can be an agent (Bonabeau
2001
) and that
its behaviour can be as simple as a reactive decision.
fl
3.5.2 Biologically-Related
As Arti
cial Intelligence tries to do, there are clear comparisons with the natural
world. Comparisons with or copying of the biological world happens often, but
trying to copy the non-biological world is less common, at least in computer
science. There are lots of processes or forces that occur in the non-biological world
that have an impact on physical systems that get modelled. Trying to integrate, or
find a more harmonious relationship between the two, could be quite an interesting
topic and computer programs might even make the non-bio processes a bit more
intelligent. It might currently have more impact in the
field of Engineering and the
paper Goel (
2013
) describes very clearly how important the biological designs are
there. With relation to a concept base, a small example of this sort of thing is
described in Sect.
6
. As noted in Wolfram (
1983
) and other places, as the second
4
Search WWH ::
Custom Search