Page 125 - 49A Field Guide to Genetic Programming

Chapter 12

Applications

Since its early beginnings, GP has produced a cornucopia of results. The
literature, which covers more than 5000 recorded uses of GP, reports an
enormous number of applications where GP has been successfully used as
an automatic programming tool, a machine learning tool or an automatic
problem-solving engine. It is impossible to list all such applications here.
In the following sections we start with a discussion of the general kinds
of problems where GP has proved successful (Section 12.1) and then re-
view a representative subset for each of the main application areas of GP
(Sections 12.2–12.11), devoting particular attention to the important areas
of symbolic regression (Section 12.2) and human-competitive results (Sec-
tion 12.3).

12.1 Where GP has Done Well

Based on the experience of numerous researchers over many years, it appears
that GP and other evolutionary computation methods have been especially
productive in areas having some or all of the following properties:
The interrelationships among the relevant variables is unknown
or poorly understood (or where it is suspected that the cur-
rent understanding may possibly be wrong). One of the partic-
ular values of GP (and other evolutionary algorithms) is in exploring
poorly understood domains. If the problem domain is well understood,
there may well be analytical tools that will provide quality solutions
without the uncertainty inherent in a stochastic search process such
as GP. GP, on the other hand, has proved successful where the appli-
cation is new or otherwise not well understood. It can help discover
which variables and operations are important; provide novel solutions

111