AI-Driven Prediction of Average Per Capita GDP: Exploring Linear and Nonlinear Statistical Techniques

Abstract

Average per capita GDP income is an important economic indicator. Economists use this term to determine the amount of progress or decline in the country's economy. It is also used to determine the order of countries and compare them with each other. Average per capita GDP income was first studied using the Time Series (Box Jenkins method), and the second is linear and non-linear regression; these methods are the most important and most commonly used statistical methods for forecasting because they are flexible and accurate in practice. The comparison is made to determine the best method between the two methods mentioned above using specific statistical criteria. The research found that the best approach is to build a model for predicting Iraq’s average GDP per capita income by relying on the amounts of average GDP per capita income in the past years (1981-2020). The researcher found that in a second way, it became clear that the non-linear regression model of the Asian model was the best model representing (average per capita GDP income) in Iraq, and this model was used to predict the period (20221-2027). When comparing the two methods of projected amounts up to 2027, it was found that the best method was the second based on the indicator mean absolute percentage error (MAPE) because he has the least value.