Thank you dude... refer it...

This is done by employing approximation accuracy measure that measures the imprecision of approximation space is employed as heuristics measure to guide the reduct process. There are two approaches for this one is forward and the other is backward feature selection. For further read 

Positive approximation: An accelerator for attribute reduction in rough set theory, Artificial Intelligence, 2010

hope this helped you, if you need further information feel free to ask