出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：In today’s changing world huge amount of data is generated and transferred frequently.Although the data is sometimes static but most commonly it is dynamic and transactional. Newdata that is being generated is getting constantly added to the old/existing data. To discover theknowledge from this incremental data, one approach is to run the algorithm repeatedly for themodified data sets which is time consuming. The paper proposes a dimension reductionalgorithm that can be applied in dynamic environment for generation of reduced attribute set asdynamic reduct. The method analyzes the new dataset, when it becomes available, and modifiesthe reduct accordingly to fit the entire dataset. The concepts of discernibility relation, attributedependency and attribute significance of Rough Set Theory are integrated for the generation ofdynamic reduct set, which not only reduces the complexity but also helps to achieve higheraccuracy of the decision system. The proposed method has been applied on few benchmarkdataset collected from the UCI repository and a dynamic reduct is computed. Experimentalresult shows the efficiency of the proposed method.
关键词：Dimension Reduction; Incremental Data; Dynamic Reduct; Rough Set Theory.
