Page 377 - Contributed Paper Session (CPS) - Volume 7
P. 377
CPS2137 Intan Mastura R. et al.
where and are complex parameters with and ℂ.
0
1
1
0
Circular statistical used in data measurement in the form of direction
and not the magnitude of the vector, where is expressed in angular size.
Both of statistical technique and statistical distribution is to analyze
random variable in form cycle there using trigonometric function.
[11] proposed a regression model to predict the mean direction of a
circular response variable from a vector of linear covariates =
( , … , ).The proposed model is given by
1
Where µ and βj are unknown parameters and ᵡj is a linear covariate
with j =1,…,p.
3. Outlier Detection in Circular Regression Model
At this time, mostly authors are performed a research based on due to the
bounded property of circular observation. Most of the paper publish today
was concentrate on detecting outlier in circular data and circular regression
model with one independent circular variable. The main aim here is to develop
an outlier detection procedure in circular regression based on 11 papers that
has been published in 2011 to 2018.
One of the methods to detecting outliers is the row deletion method. It
investigates how the deletion of any row affect the residuals, the estimated
coefficient, the estimate covariance structure of the coefficient as well as the
predicted value such DFBETAs, DFFITs and COVRATIO. In this paper, we review
some method for deleting outliers in circular regression method. The methods
are listed in table 1 with name of researcher propose and solving in short.
Author/s Ref Objective Function Proposed Optimization
Method
Alkasadi et al., 2018 [9] DFBETAs statistic Circular Regression Multiple Circular
Model Regression
Model (MCRM)
Jayant Jha and Atanu [14] MCR 1 and MCR 2 Multiple Circular – DM Circular
Biswas, 2017 Circular Regression Regression
Model Model (MCRM)
Di et al., 2017 [15] Single – linkage Down and
method Mardia Circular
– Circular
Regression
Model
Alkasadi et al., 2016 [16] COVRATIO statistic Circular Regression Multiple Circular
Model Regression
364 | I S I W S C 2 0 1 9