دراسة مقارنة لبعض طرائق تقدير النماذج اللاخطية مع تطبيق عملي

Authors

  • أ.م. د أنكين انترانيك هايك
  • مصطفى محمد كاظم

Keywords:

طريقة المربعات الصغرى، طريقة كاوس نيوتن، طريقة تعظيم دالة التوزيع اللاحق، أنموذج Negative Exponential، أنموذج Weibull.

Abstract

    The estimation of Parameters of the nonlinear models can be considered to be one of the most important things that is required in its calculation to know the response of models to the study of the data so several nonlinear models have been usedto describe the state of a certain growth Including: Negative Exponential Model and Weibull Model. The nonlinear least squares method can be considered to be the most common methods to estimate the coefficient through several formulas as in the Gauss –Newton algorithm method as well as the simplicity of this method in solving several problems. Moreover, it’s used method of Maximum a Posteriori (MAP) Estimation has also been used .

    The aim of the study is to compare the two methods of estimation using Mean squares Error (MSE) and coefficient Determination .

The study has concluded that Nonlinear least squares method is better than method of Maximum a Posteriori (MAP) Estimation because it has less MSE.

Published

2022-05-15