Rough Set Incremental Attribute Reduction Research

Attribute reduction of rough set has provided an effective tool for knowledge discovery and data analysis, and it has aroused the attention of many researchers. The so-called attribute reduction is to delete irrelevant or unimportant information while maintaining the knowledge base under the same classification ability. This is now widely utilized to get valuable and concise information.Attribute reduction is one of the core content and basis of "rough set theory", in order to obtain the necessary information in an ever-changing network information age, we use搑ough set theory to do the attribute reduction work, and many academicians have made remarkable achievements. But when confronting with the changeful database, the original method of attribute reduction becomes insufficient, this is why researching work for attribute reduction of dynamic database has begun.In this paper, two kinds of attribute reduction methods have been put forward, one is an attribute reduction for static database, which is researching with the help of relative difference comparative statement, this method is very practical, and it has provided theoretical support to learn algorithms of attribute reduction of dynamic database. The other is an attribute reduction for dynamic database, when information system target and decision-making properties is invariant while continuously increase the conditional attribute, in order to obtain the minimum reduction attribute of the system, the initial method is to recalculate all the data of the decision table, it is clear that this method is not applicable. To avoid the above situation, we should think about some ways to keep information of the original system, which only after the new system of property, on the basic of doing it, we can save a lot of time and space. Therefore, on the basis of attribute reduction of static database, the definition of relative difference comparison table has been given, and a new algorithm for minimum attribute reduction with conditional attribute has been brought forward. Through examples, we can see that this algorithm saves time and space, which result is the same with that of the traditional algorithm. Hence, the algorithm has certain applicable and referable value.