Some methods of estimating the fuzzy semi-parametric regression model with fuzzy inputs and fuzzy outputs for the Iraq Stock Exchange

Abstract

The Fuzzy Semi-Parametric Partial Linear Model is one of the important models for data analysis because it consists of two parts, a parametric and a nonparametric. The research dealt with the method of estimating the parametric part using the Fuzzy Ordinary Least Square method and estimating the nonparametric part using the Kernel Smoothing method using the functions (Triangular, Gaussian, Epanechnikov), in addition to the Cubic spline smoother method, and based on real data obtained from the website of the Iraq Stock Exchange, as the banking sector was studied in general and monthly data was taken for the period (from 1/1/2010 to 30/8/2023) and the number of views was (164) views, and with a band width of 0.5, the Speakman Gaussian method was the best estimation method because it has the lowest value of the Goodness of fit values.