On scalibility of rough set methods

We summarize some previously known scalable methods and present one of the latest scalable rough set classifiers. The proposed solution is based on the relationship between rough sets and association discovering methods, which has been described in our previous papers. In this paper, the set of decision rules satisfying the test object is generated directly from the training data set. To make it scalable, we adopted the idea of the FP-growth algorithm for frequent item-sets. The proposed method can be applied in construction of incremental rule-based classification system for stream data.