Is it possible to try to (pseudo-)optimize MPS data set with PROC GA?

[Top_Katz]

Hi!

I have some LP and MILP problems that either take a very long time to solve, or run out of memory and won't converge at all.  I am willing to trade accuracy for speed, and thought perhaps a genetic algorithm might help.  Is there a way to feed an MPS file into PROC GA?  Either directly, or perhaps someone has a macro to convert MPS into a form PROC GA can solve?

Thanks!