Building a semiparametric regression model using circular data

Abstract

Requires the use of optics that allow understanding refractive error (refractive error) and other changes with age, the human eye as an optical system for supervising visual acuity, after the refractive error is one of the main contributors to the image aid in uncorrected eyes, as a result of the presence of many phenomena in reality that provide a periodic feature only data within (0, 2π) and it is from the data that later and the analysis of these data faces. The research to build a semi-parametric regression model to enter with the nature of the circular data, the semi-parametric model-linear-cyclic was chosen for this harvest by integrating the parametric part estimated by the maximum likelihood method and assuming the random error with a joint distribution (linear-cyclic) and the estimated part using a binary circular kernel (circular kernel) different by the integrated parameter (α). Eye data were obtained from Al-Noor Specialized Eye Center in Dhi Qar Governorate using an Auto Kerato-Refractometer TOPCON TRK. 2P. The color study included 400 cases representing the right eye (OD cyl axis), the dependent eye (OD cyl axis), and the patient's age for individuals with refractive error. The results indicated that the use of the wrapped Cauchy function in the estimated part of the integrated semiparametric circular-linear-circular regression model was better than the use of the von Mises function. The researcher recommended generalizing the model to data on other phenomena.