Fuzzification of Linear Regression Models with Indicator Variables in Medical Decision Makin

Authors:
Arkady Bolotin
Affiliations:
Ben-Gurion University of the Negev, Israel
Venue:
CIMCA '05 Proceedings of the International Conference on Computational Intelligence for Modelling, Control and Automation and International Conference on Intelligent Agents, Web Technologies and Internet Commerce Vol-1 (CIMCA-IAWTIC'06) - Volume 01
Year:
2005

Citing 0
Cited 1

Classification models based on Tanaka’s fuzzy linear regression approach: The case of customer satisfaction modeling

Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology - FUZZYSS’2009

Quantified Score

Hi-index	0.00

Visualization

Abstract

To facilitate the regression analysis of the relationship between an outcome and explanatory variables in medical decision making, it is common practice to convert a continuous variable into one or more indicator variables. However, because of many uncertainties contained in medical data, linear regression models with indicator variables need modifying in order to include fuzziness. Previous studies on fuzzy linear regression analysis introduce fuzziness in the estimating models via fuzzy regression coefficients. In this study fuzziness is via the fuzzy membership functions replacing the model's indicator variables. As a result, the proposed approach does not have the common problems appearing in the usual fuzzy linear regression models.