Paper Type |
Contributed Paper |
Title |
COVRATIO Statistic for Simple Circular Regression Model |
Author |
Ali Abuzaid [a], Ibrahim Mohamed [a], Abdul G. Hussin*[b] and Adzhar Rambli [a] |
Email |
ghapor@um.edu.my |
Abstract: Few studies have considered the modeling of a linear relationship between two cir-cular variables or circular regression model. However, the problem of outlier detection in these models has not received enough consideration. This paper extends the COVRATIO statistic which is originally used to identify outliers in linear regression model. It is our aim to further exploit this approach of detecting outliers in circular regression model. The cut-off points for the statistic are obtained using simulation. It is found that the cut-off points are dependent to the sample size. We also show that the statistic has higher power of performance in detecting outliers for large sample size and large concentration parameter. A practical example is considered for illustration purposes.
|
|
Start & End Page |
321 - 330 |
Received Date |
|
Revised Date |
|
Accepted Date |
|
Full Text |
Download |
Keyword |
|
Volume |
Vol.38 No.3 (JULY 2011) |
DOI |
|
Citation |
Abuzaid A., Mohamed I., Hussin A.G. and Rambli A., COVRATIO Statistic for Simple Circular Regression Model, Chiang Mai J. Sci., 2011; 38(3): 321-330. |
SDGs |
|
View:1,612 Download:218 |