Information Technology Reference
In-Depth Information
1
Introduction:
The Biological Perspective
The aim of this chapter is to bring into focus the basic differences between
gene expression programming (GEP) and its predecessors, genetic algorithms
(GAs) and genetic programming (GP). All three algorithms belong to the
wider class of Genetic Algorithms (the use of capitals here is meant to distin-
guish this wider class from the canonical GA) as all of them use populations
of individuals, select the individuals according to fitness, and introduce ge-
netic variation using one or more genetic operators. The fundamental differ-
ence between the three algorithms resides in the nature of the individuals: in
GAs the individuals are symbolic strings of fixed length (chromosomes); in
GP the individuals are nonlinear entities of different sizes and shapes (parse
trees); and in GEP the individuals are also nonlinear entities of different
sizes and shapes (expression trees), but these complex entities are encoded
as simple strings of fixed length (chromosomes).
If we have in mind the history of life on Earth (see, for instance, Dawkins
1995 or Maynard Smith and Szathmáry 1995), we can see that the difference
between GAs and GP is only superficial: both systems use only one kind of
entity that works both as genotype and body (phenotype). These kinds of
systems are condemned to have one of two limitations: on the one hand, if
they are easy to manipulate genetically, they lose in functional complexity
(the case of GAs); on the other hand, if they exhibit a certain amount of
functional complexity, they are extremely difficult to reproduce with modifi-
cation (the case of GP).
In his topic,
River Out of Eden
, Richard Dawkins (1995) gives a list of
thresholds of any life explosion. The first is the “replicator threshold” which
consists of a self-copying system in which there is hereditary variation. Also
important is that replicators survive by virtue of their own properties. The
second threshold is the “phenotype threshold” in which replicators survive
by virtue of causal effects on something else. This “something else” is what
