2
$\begingroup$

I'm trying to do a homework problem which references "disruption" in Genetic Algorithms.

True or false? For 1-point and 2-point crossover, the schemata which have bits that are close together on the string encoding are more likely to be disrupted by crossover.

I think that it is referring to is the creation of children distinct from their parents.

For example, combining the following two parents at a single crossover point at the second digit would just create the same parents again. Thus, no disruption would occur.

p1: 00000
p2: 00001

However, combining the following two parents at a single crossover point at the second digit would create some disruption.

p1: 00000
p2: 01001
c1: 01000
c2: 00001

By intuition, I would measure disruption as the number of bits that are different in a child as compared to the two parents. However, I have no formal definitions of this, despite turning to Google. The closest I've come is to Holland's Schema Theorem, which I think is related, but I would still like a definition or an explanatory reference.

$\endgroup$
  • $\begingroup$ en.wikipedia.org/wiki/… ? $\endgroup$ – David Richerby Jun 24 '14 at 19:11
  • $\begingroup$ Okay, I guess my problem is I don't know what they mean by a schema in that context. I could really use an example. I'm going to ask my prof tonight and hopefully answer my own question by tomorrow. $\endgroup$ – Seanny123 Jun 24 '14 at 20:36
  • 1
    $\begingroup$ Ok. If you figure it out, do post the answer here so others can learn from it! $\endgroup$ – David Richerby Jun 24 '14 at 23:12
  • $\begingroup$ why dont you quote the question? "disruption" is a semiformal concept wrt GAs etc $\endgroup$ – vzn Aug 9 '14 at 17:34
  • $\begingroup$ @vzn Done. It's a badly worded question, so I wasn't sure it would help initially. $\endgroup$ – Seanny123 Aug 9 '14 at 18:57
0
$\begingroup$

According to my professor, my intuitive description of disruption is correct, however one does not measure the "degree of disruption". It is a binary measure. Either you have disruption or you do not.

Still can't find an official source for this, so edits to this answer and other answers are welcome.

$\endgroup$
  • $\begingroup$ I'm confused. Your question mentions disruption (except for one place where it mentions dispersion), but your answer mentions dispersion. Are they intended to be synonyms? What are the definitions of disruption/dispersion? I suggest you start by trying to see if you can find the definition of the concept; then you should be able to use the definition to work out the implications. $\endgroup$ – D.W. Jun 25 '14 at 19:18
  • $\begingroup$ Your professor has a link to his classes, including this one from last spring where he assigned 4 textbooks. I find it difficult to believe that you can't find an "official" source for the definition of schema, crossover, distance or disruption simply by looking in the indexes of the assigned texts. $\endgroup$ – Wandering Logic Aug 10 '14 at 17:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.