Research On Attribute And Attribute Value Reduction Method Based On Rough Sets

Z. Pawlak, Poland, has put forward Rough Sets theory that is a kind of mathematics and data analysis tool in the enghty of the 20th century. It does deal with the unaccuracy and uncertain knowledge by lower approximation and upper approxiation, and it has three kinds of ability:deducing, summing up and general knowledge reasoning. So, Rough sets theory is widely used study in the machine soon, such as knowledge acquisition, decision analysis, expert system and pattern-recognition, etc. It already becomes the focus-studied in information processing.Rough Set is a kind of tool that handles unaccurate and fuzzy knowledge. Its main thought is that if keeping categorised ability unchange, after decision reduction,it can attain the potential decision rule and potential knowledge.The main innovation of this paper is shown as follow:1. The analysis of this paper is different from the jeneral methods,it considers the reduction from the column direction.2. This paper put the attribute reduction with attribute value reduction together, while other jeneral methods first make attribute reduction, then make attribute value reduction.3. The algorithm of the paper only decision table onlye once,then computes all of the equivalence classes,then computes theirs intersections.so we can judge which attribute can be omitted and which attribute value can be deleted.