Searched over 200M research papers
10 papers analyzed
These studies suggest that various methods, including trigonometric functions, DFFITc, Satari's algorithm, graphical and numerical methods, single-linkage, circular median, and robust circular distance statistics, effectively detect multiple outliers in circular regression models.
20 papers analyzed
Circular regression models are used to study relationships between circular variables, which are common in fields such as meteorology and biology. Detecting outliers in these models is crucial as outliers can significantly affect parameter estimates and model predictions. Various methods have been proposed to identify outliers in circular regression models, each with its strengths and limitations.
Transformation-Based Methods:
DFFITc Statistic:
Clustering-Based Methods:
Graphical and Numerical Techniques:
Robust Circular Distance:
Non-Parametric Methods:
Outlier detection in circular regression models is a well-researched area with various effective methods available. Transformation-based methods, DFFITc statistic, clustering-based methods, graphical and numerical techniques, robust circular distance, and non-parametric methods all show promise in identifying outliers. Each method has its strengths, and the choice of method may depend on the specific characteristics of the data and the model.
Most relevant research papers on this topic