Variable Precision Fuzzy Rough Model for Data Analysis

A variable precision fuzzy rough data model(VPFRDM) that is an extension of Ziarko’s variable precision rough set(VPRS) model was proposed based on fuzzy set.A method for attribute reduction from a fuzzy decision information system was developed,in which the attribute importance is taken as the heuristic information.Fuzzy knowledge is extracted through calculating classification ability of each fuzzy pattern class to the decision categories.Simulation results show that the VPFRDM is effective and has better data generalization ability compared with Ziarko’s VPRS.