A quick incremental updating algorithm for computing core attributes

Abstract

Computing core attributes is one of key problems of rough set theory. Many researchers proposed lots of algorithms for computing core. Unfortunately, most of them are designed for static databases. However, many real datasets are dynamic. In this paper, a quick incremental updating core algorithm is proposed, which only allies on the updating parts of discernibility matrix and does not need to store, re-compute and re-analyze discernibility matrix, when new objects are added to decision table. Both of theoretical analysis and experimental results show that the algorithm is effective and efficient. © 2010 Springer-Verlag Berlin Heidelberg.