A Cache-Based Attribute Granular Computing System

Abstract

This paper presents a multilayer cache-based attribute granular computing system. After defining an attribute granule, it brings forward an AHP based attribute granules combination algorithm which can be used for dimensionality reduction in the preprocess stage of data mining. A cache-based operation and storage model is presented then. The storage management of the attribute granules is discussed mainly. Finally a hierarchy cache-based attribute granular computing system is presented. It can be used to implement the AHP based attribute granules combination algorithm and other algorithms based on attribute granular computing.