Rough Set Theory Basic Algorithm Research And Application In Knowledge Discovery

With the development and application with the computer and the large-scale storage system, the massive data was accumulated by many industries and companies. While process these data, traditional data analysis and retrieval method, not only will cost the massive computing time, moreover depend on the data relations supposition and the estimate. So it has urged people to develop a set of practical techniques and methods for intelligently and automatically drawing the interesting knowledge from a very large number of data having been continuously generated. Early in 1980s, a Polish mathematician Pawlak.Z advocated the Rough set theory. It was a new mathematical tool for processing vagueness and uncertainty. It can analyze the facts hidden in the data without any additional knowledge about the data.This article discusses the Rough set theory application in knowledge discovery. The work mainly focuses on two things, attribute reduction and rough query. The main works were listed as follows:(1) An attributes reduction algorithm based condition attributes'classification power is proposed and discussed.(2) Propose and prove an equivalent and efficient method for computing positive region.(3) Research the incremental computing of positive region; and deduces some lemma to decrease the numbers of objects in the source decision table. Then an attribute reduction algorithm is proposed and its completion is proved. The algorithm is tested by the data sets from the machine-learning database, UCI.(4) Analysis the knowledge discovery system process. Propose a method of rough query. By supplying a real value, this method is useful to deal the rough query for different purposes.