Using ReliefF to predict the class of the next number returned by PRNG.

I'd like to duplicate the results achieved in this paper http://lkm.fri.uni-lj.si/rmarko/papers/Savicky08-AAI.pdf (page 9) using ReliefF to predict the class of the next number given to me by a PRNG. The problem is I'm completely new to R. I see that in the Core package there's a function that will "evaluate attributes" using ReliefF, but after attributes have been evaluated, how do I use this to predict the next class? Further, how do I incorporate what I've received into the model so that I can predict the n+2 class?

Working code would be greatly appreciated.