Parameter Estimation by Anfis where Dependent Variable has Outlier
Yazarlar (3)
Karadeniz Teknik Üniversitesi, Türkiye
Prof. Dr. Kamile ŞANLI KULA
Kırşehir Ahi Evran Üniversitesi, Türkiye
Ankara Üniversitesi, Türkiye
| Makale Türü |
Özgün Makale
(SSCI, AHCI, SCI, SCI-Exp dergilerinde yayınlanan tam makale)
|
||
| Dergi Adı | Hacettepe Journal of Mathematics and Statistics | ||
| Dergi ISSN | 1303-5010 | ||
| Dergi Tarandığı Indeksler | SCI-Expanded | ||
| Makale Dili | İngilizce | Basım Tarihi | 04-2014 |
| Cilt / Sayı / Sayfa | 43 / 2 / 309–322 | DOI | – |
| Özet |
| Regression analysis is investigation the relation between dependent andindependent variables. And, the degree and functional shape of this relation is determinate by regression analysis. In case that dependentvariable has outlier, the robust regression methods are proposed tomake smaller the effect of the outlier on the parameter estimates. Inthis study, an algorithm has been suggested to define the unknownparameters of regression model, which is based on ANFIS (AdaptiveNetwork based Fuzzy Inference System). The proposed algorithm, expressed the relation between the dependent and independent variablesby more than one model and the estimated values are obtained byconnected this model via ANFIS. In the solving process, the proposedmethod is not to be affected the outliers which are to exist in dependentvariable. So, to test the activity of the proposed algorithm, estimatedvalues obtained from this algorithm and some robust methods are compared. |
| Anahtar Kelimeler |
| Adaptive network | fuzzy inference | robust regression |
