I thought about it,
and now I see, that in most real-world problems the "ideal" solution cannot
be specified, because no one knows,
if it is really the best possible solution, that can exist.

Now I see, that the need for stopping condition when the "right" solution
was evolved, was born in world of academic examples.
When using GA in real use-cases, this is very rare requirement.

Thanx for good question ;)
Best regards
Stefan






On Sun, Aug 25, 2013 at 4:33 PM, Gilles <[email protected]>wrote:

> On Sun, 25 Aug 2013 01:30:11 +0200, Štefan Šimík wrote:
>
>> I checked the genetic algorithm - and it works so that
>> it evolves new generations until stopping condition returns true.
>>
>> There are 2 implementations of StoppingCondition now:
>> 1) FixedElapsedTime
>> 2) FixedGenerationTime
>>
>> What I am missing in the algorithm is the concept of automatic finishing,
>> when ideal/optimal solution was found.
>> I think, this should be implemented in the basic algorithm without
>> additional programming (such as implementing custom
>> StoppingCondition, which stops algorithm when ideal solution was found).
>>
>> What do you think about it guys?
>>
>
> How does one know when the "ideal" solution is found?
>
> Regards,
> Gilles
>
>
>
> ------------------------------**------------------------------**---------
> To unsubscribe, e-mail: 
> user-unsubscribe@commons.**apache.org<[email protected]>
> For additional commands, e-mail: [email protected]
>
>

Reply via email to