e genetic algorithm

as developed based on the genetic algorithm (GA), it is better to

troduce GA first and introduce the basic operators used in GA

he main GA operators are also used in GP.

A, a combinatory problem is expressed by a chromosome, which

g. If a combinatory problem is binary, the chromosome or the

r a problem is a binary vector. Suppose there are ten variables

⋯, ݔଵ଴) as the potential contributors to formulate a system. Before

ng a system, it is unknown whether all these ten variables must

yed or part of them can be employed. Suppose this ten-variable

expressed by a function shown below, which is either an analytic

or a complex function,

ݕൌ݂ሺݔଵ, ݔଶ, ⋯, ݔଵ଴ሻ

termine which of these ten variables should finally contribute to

of a practical problem, a string of presence/absence of variables

ed. Such a string is called a chromosome. Figure 8.3 shows two

omes of two potential solutions to this problem, where the upper

ws that four variables ݔଷ, ݔହ, ݔ଼ and ݔଵ଴ are employed to define

m. The lower panel shows that three variables ݔଷ, ݔସ and ݔଵ଴ are

d to define the system. The upper panel of Figure 8.3 will end up

odel expressed as ݂ሺݔଷ, ݔହ, ݔ଼, ݔଵ଴ሻ and the lower panel of Figure

esents a model expressed as ݂ሺݔଶ, ݔସ, ݔ଻ሻ. Both are more

ious than the use of ten variables. However, which has the best

a system is unknown and needs an evaluation.

wo chromosome expressions of ten variables. A ‘1’ in either chromosome stands

ence of a variable and a ‘0’ stands for the absence of a variable.