A fuzzy logistic regression model based on the least squares estimation
- University of Shanghai for Science and Technology (China)
To construct the relationship between fuzzy inputs and fuzzy output described by linguistic variable, we develop a fuzzy logistic regression model, which can be applied for various problems, such as clinical research, risk investment and decision making. In this regard, we introduce the integral calculation of distance of cut sets and a fuzzy adjustment term which could prevent a large fuzzy error for a fuzzy output caused by the output degenerating into crisp number when the independent variables are crisp numbers. The parameters of the fuzzy logistic regression model are derived by means of the least squares method. We adopt three criteria involving mean square error, Kim and Bishu criterion and capability index and also perform two ways of prediction including the in-sample forecast as well as the one-leave-out cross validation. The comparisons with five existing methods show that our proposed method has satisfactory performance and the results are illustrated in some case studies.
- OSTI ID:
- 22769269
- Journal Information:
- Computational and Applied Mathematics, Journal Name: Computational and Applied Mathematics Journal Issue: 3 Vol. 37; ISSN 0101-8205
- Country of Publication:
- United States
- Language:
- English
Similar Records
Application of fuzzy system theory in addressing the presence of uncertainties
The fuzzy regression approach to peak load estimation in power distribution systems