|History|| Rough uncertain concept that set catch, and grey model it closes to be different with fuzzy logic. Whom fuzzy cognition that logic catch have uncertain, grey model deal with little uncertainty that sample lead to the fact, namely the incomplete (information incompleteness ) prediction that put of information, as for assembling coarsly it is an inconsistency(inconsistency ) used for dealing with the materials.
The Rough Set was proposed by Zdzislaw Pawlak in 1982, the groundwork of this theory was to attempt to use the the least characteristic attribute (attributes ), divide the subject matter (objects ) with the same decision attribute (decision attributes ) into the same class, reduce the proportion of subject matter that can`t be distinguished as much as possible , namely is it approach accuracy(accuracy of approximation ) to improve, most succinct in order to to form and effective classification predict rule(rule). It is different on the method in the fuzzy and logic characteristic , it needn`t any external parameter, is it get give message of the materials definitely to use to need only.
The Rough Set has already been received in the economic and financial educational circles, extract for combining the automatic rule (automation of rule extraction) scholar of the materials mining that judges (domain expertise )with the expert with interested technology, the attention of the height. Slowinski and Zopounidis (1995 ) uses the coarse set method to carry on the assessment of the bankrupt risk. Skalkoz (1996 ) uses the technology of the coarse set to extract the rule of a group of short-term trade OEX index. Hampton (1997 , 1998a , 1998b ) and Mozek and Skabek (1998 ) offer the detailed explanation of the economic and financial scholar`s coarse set theory . And include the article of Doumpos and Zopounidis even more in one book of Chen and Wang (2003 ), it is to being coarse for set it is and not traditional for statistics for tool not categorised that have (for instance: Discriminate analysis,logit and probit analysis) categorised performance last tight one comparison.