GENE EXPRESSION PROGRAMMING: in solving friction factor in turbulent flow

Genetics has always amazed me ever since we have it in my high school days. The gene expression process is simple in what I remember from my bio teacher's lectures. But in reality, this process is so complex because from the genetic code stored in the DNA, it will transform series of codes into gene product in the form of proteins. Then, these proteins will in turn form the phenotype or the evident characteristics of an organism. The ingenuity of gene expression in organisms has been imitated to create the so called "gene expression programming".  The mechanism of process is the same but instead of organisms, gene expression programming produces algorithms or models that can change and adapt to a certain environment. 

GEP has the same genotype-phenotype system just like in a living organism. It depends on genomes composed of numbers which represent strings of symbols and whose symbols further represent equations (which is analogous to phenotypes). The genome can be illustrated by a binary tree from which you can trace its nodes to evaluate the equation. This is a very powerful technique in programming because you are not mapping hard values but symbols instead.  

Well, I think you have to try to use it first to create a program in order to fully understand how it works. This has been a new knowleadge for me but GEP and their use actually dates back to the 1950s where they were first used to solve optimization problems (e.g. Box 1957^[1] and Friedman 1959^[2]). But it was in 1965 when this evolutionary technique started to gained popularity.  Nowadays, GEP has already a lot of applications especially in different engineering branches.  

In the field of chemical engineering, GEP has been used to model some highfalutin empirical equations related to fluid mechanics.  One example, in hydraulic design , GEP has been used as a technique to estimate the friction factor using the Colebrook–White equation for turbulent fluid-flow.This is very important for scientific intensive computations because numerical simulations of pipe flows require the computation of the friction coefficient for each point. And in simulations of long pipes, the friction coefficient has to be solved many times. GEP makes the calculation faster and more accurate. 

The Colebrook–White equation:

*This conclusion is based from a journal I have read in science direct. The paper basically examines the potential of genetic programming based technique in estimating flow friction factor in comparison with the most currently available explicit alternatives to the Colebrook–White equation.