The genetic algorithm and its possible applications

[RomFil]

Good afternoon!

Came across this article https://www.forextimes.ru/foreks-stati/geneticheskie-algoritmy-v-poiskax-svyashhennogo-graalya (admins know that I violate, but the article is from 2005 ..., please do not punish strongly).

Now a question - do you think the proposed method contains any rational arguments? Basically what is a genetic algorithm and its use to solve optimization problems I know firsthand. I want, if there are willing, to try to repeat, but need ideas on the description of the possible inputs and outputs, but taking into account the possibility of an objective description. Of course, I can "invent" these regularities myself, but I don't have time yet. Maybe there are already ready "simple" laws. Thank you.

Regards, RomFil

[Реter Konow]

RomFil:

Good afternoon!

Came across this article https://www.forextimes.ru/foreks-stati/geneticheskie-algoritmy-v-poiskax-svyashhennogo-graalya (admins know that I violate, but the article is from 2005 ..., please do not punish strongly).

Now a question - do you think there is a rational concept in the suggested method? Basically what is a genetic algorithm and its use to solve optimization problems I know firsthand. I want, if there are willing, to try to repeat, but need ideas on the description of the possible inputs and outputs, but taking into account the possibility of an objective description. Of course, I can "invent" these regularities myself, but I don't have time yet. Maybe there are already ready "simple" laws. Thank you.

Regards, RomFil

Interesting article. It is very well explained. But the author seems to confuse system "parameters" and "values" of those parameters. This creates some inaccuracy in expressions. If one individual is an assembly of system parameters, then the reproduction of individuals, is the "reproduction" of parameter values, not the parameters themselves. And it is the values we are looking for, not the parameters. The parameters are the system, and the system is unchanging as we "genetically" search through the values of the "genes".

The approach has a huge future in the field of automatic design of complex systems. Nature, after all, has "bred" them that way.

[RomFil]

Реter Konow:
Interesting article. It is very well explained. But the author seems to confuse the 'parameters' of the system with the 'values' of those parameters. This creates some imprecision in expressions. If one individual is an assembly of system parameters, then the reproduction of individuals is the "reproduction" of parameter values, not the parameters themselves. And it is the values we are looking for, not the parameters. The parameters are the system, and the system is unchanging as we "genetically" search through the values of the "genes".

I agree about "confusion". Rather it seems to me that the author understands everything, but has stated it in a slightly different way, so there is a mess, but after all it is not the main thing ... The main thing is the approach.

[RomFil]

Реter Konow:
Interesting article. Explained very clearly. But the author seems to confuse the 'parameters' of the system with the 'values' of those parameters. This creates some imprecision in expressions. If one individual is an assembly of system parameters, then the reproduction of individuals is the "reproduction" of parameter values, not the parameters themselves. And it is the values we are looking for, not the parameters. The parameters are the system, and the system is unchanging as we "genetically" search through the values of the "genes".

The approach has an enormous future in the field of automatic design of complex systems. Nature, after all, has "bred" them that way.

It's been 15 years ... :) And you think people don't still use this approach? The "future" for this method is probably past. :)

[Реter Konow]

RomFil:

It's been 15 years ... :) And you think people don't still use this approach? The "future" for this method is probably past. :)

This approach by itself is not enough. It can serve as a tool for "polishing" the ready systems - checking and debugging.

You can't build systems like that. You need a different algorithm. When it is invented, GA will be remembered.

[Vladimir Perervenko]

RomFil:

Good afternoon!

Came across this article https://www.forextimes.ru/foreks-stati/geneticheskie-algoritmy-v-poiskax-svyashhennogo-graalya (admins know I'm breaking it, but the article is from 2005..., please don't punish me too much).

Now a question - do you think there is a rational concept in the proposed method? Basically what is a genetic algorithm and its use to solve optimization problems I know firsthand. I want, if there are willing, to try to repeat, but need ideas on the description of the possible inputs and outputs, but taking into account the possibility of an objective description. Of course, I can "invent" these regularities myself, but I don't have time yet. Maybe there are already ready "simple" laws. Thank you.

Regards, RomFil

How hard can it be to search the site?

[RomFil]

Vladimir Perervenko:

Is it difficult to search the website?

Don't worry about me ... :) I will... "search" ... :)

[Dmitry Fedoseev]

RomFil:
Don't worry about me ... :) I will... "search" ... :)

And look in the terminal, in the tester, it has a genetic optimisation function.

[RomFil]

Dmitry Fedoseev:

And look in the terminal, in the tester, it has a genetic optimisation function.

Don't worry either ... :) I know. I use it. But I don't like it very much... :) Even for the sole reason of not being able to make my target, it doesn't suit me. Thanks.

[RomFil]

Here's the guinea pig ... :) https://www.mql5.com/ru/forum/335027

[Dmitry Fedoseev]

RomFil:
Don't worry either ... :) I know. I use it. But I don't like it very much... :) Even for the sole reason of not being able to make my target, it doesn't suit me. Thanks.

https://www.mql5.com/ru/search#!keyword= own%20criteria%20optimisation