Using the Circular Distance Principle to Detect Outliers in the Circular Regression Model

Abstract

In order to detect abnormal values in the circular data that represent realistic data about the eye, which is an angle, the COVA RATIO statistic was calculated in this study using the principle of circular distance, after the abnormal data was removed from the entire data set and consideration of the covariance determinant of the covariance matrix. The amblyopia-causing eye in order to estimate the data's circular regression model. It was determined that the application of the contrast ratio statistics helped in discovering the abnormal values that affect the eye completely, and whose format differs from the rest of the measurements for patients, as the circular regression model provided a good fit for the real data in light of the abnormal observations.