.png){kind=icon}
Return
Fuzzy regression analysis based on M-estimates
.png){kind=icon}
PRE
AI
DOI：10.1016/j.eswa.2021.115891.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The least-squares technique has been shown to possess valuable properties as a method of the parameter estimation of classic and fuzzy regression analysis. However, the behavior and properties of the least-squares estimators are affected when outliers arise in the sample and/or by slight changes in the dataset. Robust techniques, on the other hand, provide robust estimators of the parameters which avoid such adverse effects. For this purpose, this paper extends the M-estimation approach to fuzzy regression analysis which provides consistent results in the presence of outliers. The parameters estimation problem is reduced to a reweighted algorithm which is a simple approach both theoretically and computationally. The proposed algorithm decreases the effect of outliers on the model fit by down-weighting them. To show the performances of the proposed method against some commonly used fuzzy regression models simulation studies, and two applicative examples based on real-world datasets in hydrology and atmospheric environment are provided. The sensitivity analysis of the estimated parameters are also reported based on a Monte-Carlo simulation study showing the efficiency of the proposed estimators in comparison with some other well-known methods in fuzzy regression analysis.
Keywords:
Fuzzy outlier
Goodness-of-fit
Huber function
Reweighted algorithm
Robustness
Journal
IF:
7.5
Papers:
3.0W
Citations:
10.2W
Organization
View more organizations
Cited Papers
ShareSaveEstimation of high dimensional mean regression in the absence of symmetry and light tail assumptions
OAAI
OAAIShareSaveShareSaveAffordable levels of house prices using fuzzy linear regression analysis: the case of Shanghai.png)
SOFT COMPUTING
SOFT COMPUTINGIF2.5
PREAI
ShareSave View more