Page 375 - Contributed Paper Session (CPS) - Volume 7
P. 375
CPS2137 Intan Mastura R. et al.
Detecting outlier in a circular regression model –
A review
Intan Mastura Ramlee, Safwati Ibrahim
Institute of Engineering Mathematics, University Malaysia Perlis, Pauh Putra Main Campus,
02600 Arau, Perlis
Abstract
Presently, circular data is very relevant and important application technique in
many fields such as Biology, Medicine and others. Whereas one type of data
direction is a data circular. In this case, author have a tendency to study and
explore in detail about circular regression model. In this paper aim to review
the outlier detection methodologies in circular regression model based on 11
articles. In general, this paper performs a survey of circular regression from
2011 to 2018 in order to see the trend of current study. Here, we concentrate
the attention on the methodologies of identifying outlier in this model. This
survey of circular regression model in which many interesting properties and
is good enough to detect the occurrence of outlier. Through the survey may
highlight the significant of methodologies to detect outliers in circular
regression model and provide guideline for future work to look into the
research gap.
Keywords
Circular regression model, outlier detection, statistical analysis
1. Introduction
Statistical analysis is the study the relationship between the independent
variable and dependent variable in regression analysis. Research on circular
variable in regression model has long tradition since four decade ago. The
field of the circular regression is referred of relationship when both
explanatory and response variable are circular. Circular data arise in many
different fields such as biology, physics, geology, medical science and others.
Thus, a practical need for circular regression can been in real-life problem such
the wind direction and the direction of movement of clouds, the arrival of
patient (24 hours) in the emergency room in a hospital and others [1].
Data circular is types of data direction whereas the data measurement take
range in degree direction or unit time [2]. The analysis of circular data is the
measurement value of data are repeated periodically. Therefore, the circular
data inaccurate to analyse with linear statistic, so that circular data needs to
be analysed by using circular statistics. As a result, the statistical software have
been provided including Axis, DDSTP, Oriana, MATLAB and R/S language for
circular data analysis.
362 | I S I W S C 2 0 1 9