Details regarding the genetic code and crossover in the Genetic Algorithm.

I am trying to understand some details within the GA optimizer in the Global Optimization Toolbox. If "double-precision vector" is chosen to define the population, what is the "genetic code" used by GA - is it 64-bit binary for each variable? Also, does crossover apply to the individual genes (e.g., each bit of the binary string) or does it / can it apply to alleles (i.e., groups of genes such as one complete variable). Thanks!!!

Answer 1

[Alan Weiss]

You can find most of the details in the documentation:

<http://www.mathworks.com/help/gads/how-the-genetic-algorithm-works.html>

<http://www.mathworks.com/help/gads/genetic-algorithm-options.html#f7820>

The basic thing to know is that GA is not working on bits, it is trying to optimize a real-valued function of real variables.

If your population type is "bitstring", where each component of the decision variable vector is a 0 or 1, then look at this section of the documentation.

Alan Weiss

MATLAB mathematical toolbox documentation