Combine maximum entropy with Fourier series residual to estimate the modified exponential growth model

Saifaldin Hashim Kamar; Hala Kadhum Obead; Aseel Mahmood Shakir

Saifaldin Hashim Kamar Al-Iraqia University, College of Management and Economics.
Hala Kadhum Obead Al-Iraqia University, College of Management and Economics.
Aseel Mahmood Shakir Al-Iraqia University, College of Management and Economics.

Abstract

The objective of this paper is to estimate the Modified Exponential Growth model using Maximum Entropy method compact with Fourier Series Residual, and applied this model to review the gap between the actual and expected prices of OPEC basket of crude oil.\\ Method: This paper presented the method of Maximum Entropy using Fourier Series Residual. The first step was generated four samples of size (10, 15, 20, 25) based on modified exponential growth model using the simulation technique. The second step compared the fitting modified exponential growth model using methods of Maximum Entropy $(ME)$ and Modified Maximum Entropy $(MME)$ depending on Mean Absolute Percentage Error $(MAPE)$ and Root Mean Squares Error $(RMSE)$. We applied $(ME)$ and $(MME)$ on real data selected from the prices of crude oil for the period from June 2018 to June 2019.\\ Result: According to each method used and for the four generated models, the result showed that the best method is Modified Maximum Entropy $(MME)$.