Possibilistic data analysis and its similarity to rough sets

Authors:
Hideo Tanaka;Peijun Guo
Affiliations:
Graduate School of Management and Information, Toyohashi Sozo College, Toyohashi-shi 440-8511, Japan;Faculty of Economics, Kagawa University, Takamatsu, Kagawa 760-8523, Japan
Venue:
Data mining, rough sets and granular computing
Year:
2002

Citing 4
Cited 1

Rough classification

International Journal of Man-Machine Studies
Identification of possibilistic linear systems by quadratic membership functions of fuzzy parameters

Fuzzy Sets and Systems
Portfolio selection based on fuzzy probabilities and possibility distributions

Fuzzy Sets and Systems
Interval regression analysis by quadratic programming approach

IEEE Transactions on Fuzzy Systems

Dual models for possibilistic regression analysis

Computational Statistics & Data Analysis

Quantified Score

Hi-index	0.02

Visualization

Abstract

This paper is dealing with the upper and lower approximation models for representing the given phenomenon in a fuzzy environment. Based on the given data, the upper and lower approximation models can be derived from upper and lower directions, respectively where the inclusion relationship between these two models holds. Thus, the inherent fuzziness existing in the given phenomenon can be represented by the upper and lower models. The modalities of the upper and lower models have been illustrated in regression analysis and also in the identification methods of possibility distributions. The comparison of the concepts of possibility data analysis and rough sets is shown. A measure similar to the accuracy measure of rough sets is used to clarify the difference between the data structure and the assumed model.