The Study, Based On The Certainty Of S-rough Sets Decision Rules

Rough sets theory is a mathematic tool to deal with fuzzy and uncertain knowledge,which gives study on equivalence class idea and portrays set by its lower approximation and upper approximation. Rough sets theory is put forward by Poland mathematician Z. Pawlak in 1982.From then on,its theory and application study have been extensively developed especially in data mining field.Rough sets theory offers us a convenient tool to deal with great capacity for liquormagnanimity data produced by current information society. Professor SHI Kaiquan extended Z.Pawlak rough sets into singular rough sets, called S-rough sets for short in 2002,which introducestransfer functions F = {f₁,f₂,∧,f_n} and F' = {f₁',f₂',∧,f_m'}based on classic roughsets. Singular rough sets makes the study spans from static fields to dynamic fields.This broadly expands the theory study and application fields of rough sets theory. Then Professor SHI Kaiquan put forward variation rough sets and variation S-rough sets.In which professor SHI explain rough sets essentially by attribute classes and element classes dual.We can educe the decision regulation and classify regulation of the problem through attribute reduction in rough sets theory, retaining classify capacity. Uncertain decision rules are not welcome to us in decision rule. Illumed by variation S-rough sets and knowledge filtration idea,we combine variation S-rough sets and the produceof decision rules. We introduce transfer functions F = {f₁,f₂,∧,f_n} andF'={f₁',f₂',∧,f_m'} to condition attributes, and we input and output attributes tocondition attributes. We give discussion on the influence of one direction input,one direction output and two direction input and output to condition attributes. And we obtain one direction input,one direction output and two direction input and output theorem. Thereby ,uncertain decision rules can be filtrated to certain decision rules. The trouble brought by uncertain decision rules can be eliminated.