| Research Title |
A modified GM-estimation for Robust Fitting of Mixture Regression Models |
| Date of Distribution |
12 August 2014 |
| Conference |
| Title of the Conference |
The 2nd ISM International Statistical Conference 2014 with Applications in Sciences and Engineering |
| Organiser |
Universiti Malasia PAHANG |
| Conference Place |
M.S. Garden Hotel |
| Province/State |
Kuantan/ Pahang |
| Conference Date |
12 August 2014 |
| To |
14 August 2014 |
| Proceeding Paper |
| Volume |
1643 |
| Issue |
- |
| Page |
264 - 269 |
| Editors/edition/publisher |
Editors Mohd Sham Mohamad, Wan Nur Syahidah Wan Yusoff, Nor Aida Zuraimi Md Noar, Roslinazairimah Zakaria and Mohd Rashid Ab Hamid |
| Abstract |
In the mixture regression models, the regression parameters are estimated by maximum likelihood estimation (MLE) via EM algorithm. Generally, maximum likelihood estimation is sensitive to outliers and heavy tailed error distribution. The robust method, M-estimation can handle outliers existing on dependent variable only for estimating regression coefficients in regression models. Moreover, GM-estimation can handle outliers existing on dependent variable and independent variables. In this study, the modified GM-estimations for estimating the regression coefficients in the mixture regression models are proposed. A Monte Carlo simulation is used to evaluate the efficiency of the proposed methods. The results show that the proposed modified GM-estimations approximate to MLE when there are no outliers and the error is normally distributed. Furthermore, our proposed methods have more efficient than the MLE, when there are leverage points. |
| Author |
|
| Peer Review Status |
ไม่มีผู้ประเมินอิสระ |
| Level of Conference |
นานาชาติ |
| Type of Proceeding |
Full paper |
| Type of Presentation |
Oral |
| Part of thesis |
true |
| ใช้สำหรับสำเร็จการศึกษา |
ไม่เป็น |
| Presentation awarding |
false |
| Attach file |
|
| Citation |
0
|
|
|
Publication
Journal Publication
ไทย