Similar To Rough Sets And Knowledge Discovery

The rough set theory was put forward by Professor Z.Pawlak, Poland Mathematician, as a math theory to study uncertain knowledge. It can analysis and process effectively all kinds of immaturity information as inexactness, uncertain, uncompleted, find implicative knowledge and open out potential rules. Indiscernibility relations is its basic thought, when the object attribute are very many and with the value precise expression, according to the rough set definition, two objects attribute value must be completely same, just have equivalent relations between them.namely indiscernibility relations. Because it was easy to receive the small data difference the influence, this enormously limited widespread and the accuracy which the rule extract. When people know objective world and analysis solves in the question process, it have the massive similar problems. This article spread the classics rough set theory based on this kind of thought, using the similar relations replaced equivalent relations in the rough set. This time the similar relation is no longer forms to the division that gathers, between them are mutually overlapped. The practice proves, the similar model is better in the use compared to the classics rough set. This article produced under the similar relational rough collection model, studied their some basic property, and how discussed utilizes the similar relations discovers question and so on knowledge as well as attribute reduction.The collection value information system is also one kind of similarrelations, it is the incomplete information system, the traditional relational database does not support the unit containing the collection value, but in the reality, the collection value information system is inevitable, this time we cannot obtain some information easily. The traditional rough set research object is the complete information system, it can effectively use reduction algorithm analysis decision table, withdraws latent pattern which in the policy-making table between the attribute exists. Many scholars conducting the thorough research on null values information system, obtained some very good knowledge-gained method, but stepped in under the collection value the rough set research are very few. Finally, the article also produced in the collection value information similar relations in the processing incomplete information utilization, and derived its recursion formula.