Sorry...I have some problem...

[Aleksey Smirnov]

Hi, Park !

 P> I have objective data, and a number of data
 P> is 900.
 P> I have optimized this. The method that I used
 P> was mapping 900 to 1024, duplicating data.

There are several ways of remapping: fixed, random, and probabilistic.
In fixed remapping a redundant value is remapped to a fixed valid value.
In random remapping a redundant value is remapped to a valid value at random.
This avoids the representational bias problem, but also causes less information
to be passed on from parents to children.
In probabilistic remapping every gene value is remapped to one of two valid
values in a probabilistic way. Thus each valid value is equally likely to be
represented.

Regards,
        Aleksey