Grid-clustered rough set model for self-learning and fast reduction

Rough set theory has been playing a significant role in data mining, and its progress to intelligentization in future requires more abilities, such as hybrid data processing, fast attribute reduction and self-learning. The essential demand of the three abilities is the knowledge depicting of the considered universe. Grid subspace cluster(GSC) algorithm characterized by densities and distances in grid subspace is presented along with the automatically selecting of cluster centers, which is regarded as the knowledge depiction in rough set model, i.e. grid-clustered rough set(GCRS) model. For the ability of self-learning, rough self-learning theory including extensional learning and intensional learning is raised. Subsequently, a fast attribute reduction algorithm and a rough self-learning algorithm based on GSC, rough self-learning theory and GCRS model are designed. A multitude of experiments substantiate that, GCRS model could meet the future demands of rough set theory.147