The model of fuzzy variable precision rough sets

  • Authors:

  • Suyun Zhao;Eric C. C. Tsang;Degang Chen

  • Affiliations:

  • Department of Computing, Hong Kong Polytechnic University, Kowloon, Hong Kong and School of Mathematics and Computer Sciences, Hebei University, Baoding, China;Department of Computing, Hong Kong Polytechnic University, Kowloon, Hong Kong;Department of Mathematics and Physics, North China Electric Power University, Beijing, China

  • Venue:
  • IEEE Transactions on Fuzzy Systems
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

The fuzzy rough set (FRS) model has been introduced to handle databases with real values. However, FRS was sensitive to misclassification and perturbation (heremisclassificationmeans error or missing values in classification, and perturbation means small changes of numerical data). The variable precision rough sets (VPRSs) model was introduced to handle databases with misclassification. However, it could not effectively handle the real-valued datasets. Now, it is valuable from theoretical and practical aspects to combine FRS and VPRS so that a powerful tool, which not only can handle numerical data but also is less sensitive to misclassification and perturbation, can be developed. In this paper, we set up a model named fuzzy VPRSs (FVPRSs) by combining FRS and VPRS with the goal of making FRS a special case. First, we study the knowledge representation ways of FRS and VPRS, and then, propose the set approximation operators of FVPRS. Second, we employ the discernibility matrix approach to investigate the structure of attribute reductions in FVPRS and develop an algorithm to find all reductions. Third, in order to overcome the NP-complete problem of finding all reductions, we develop some fast heuristic algorithms to obtain one near-optimal attribute reduction. Finally, we compare FVPRS with RS, FRS, and several flexible RS-based approaches with respect to misclassification and perturbation. The experimental comparisons show the feasibility and effectiveness of FVPRS.