Study And Application Of Reduction Algorithms Based Rough Set In Knowledge Discovery

The work of this present paper mainly includes the investigation of the structure of KDD system, the study of the basic concepts of rough set theory, the research of the questions in the heuristic reduct finding algorithms based the rough set theory used in the data reduction and the discretization in the field of KDD, and the development of the system which is used to study the algorithms of the rough set.Through the study in the data reduction, this paper presents a new reduction algorithm named HORAFA-SVDM and an incremental algorithm of the heuristic optimal reduct finding algorithm of the frequencies function named HORAFAæ¡°A. According to the results of the experiment, the paper discusses the question that it takes too much time in the finding reduct when this kind of the algorithm is used in the data set without core. An adaptive model HORAFA-A is constructed with the heuristic optimal reduct finding algorithm presented in the paper.In the field of discretization, frequencies function is imported to the NS algorithmand the new reduct finding algorithm are shown. A new concept ------Candidate Coreand a new algorithm based on it are presented to solve the question in the NS algorithm with the heuristic reduct finding algorithm.At last, throgh the study of some prototype systems of KDD, this paper introducesan experiment system------RSDMES built in the research work, which is used to studythe algorithm of rough set.