Re: [cilib-devel] Higher level response strategy for dynamic MOO problems

[Gary P. <gpa...@gm...>]

Hi Marde,

I was actually wondering when we would need functionality as you have 
described here. The current detection / response implementation is a little 
limited as you have discovered.

I'll have a look at it. The potential solution I'm thinking of might require a 
fair amount of work to get it going.

Regards,
Gary

On Tuesday 02 June 2009 23:14:17 Marde Greeff wrote:
> Hi everyone,
>
> I want to do the following:
>
> For the VEPSO algorithm that solves a dynamic MOO problem, the following
> must be possible:
>
> 1) All swarms detect for a change
> 2) If any swarm detects a change, all swarms respond to the change
>
> However, with the current implementation the detection and response occurs
> on the swarm level in the DynamicIterationStrategy class.
>
> What is the best way to do this?
>
> Marde