بناء نموذج قياسي للاستيراد في السودان باستخدام الانحدار المتعدد و الشبكات العصبية الاصطناعية

الملخص EN

This study dealt with building econometric model for import in Sudan using multiple regression and artificial neural networks.

Besides it aimed at finding variables and factors that affect the import in Sudan by using multiple regression technique.

Unite root test and co-integration were applied to the data and by taking advantage of artificial neural networks technique, used in quality assurance of the econometric model.

The study relied on statistical analysis methodology and econometric approach.

By using multiple regression, the findings of the study revealed that there was a significant effect on import in Sudan due to the variables: GDP, investment, exchange rate, and duties.

The neural networks have shown that the most influential variables on import were: investment, GDP, and duties (respectively).

Besides, the findings illustrated an effect to a lesser extent due to the variables: the exchange rate and exports.

However, the two variables: consumer spending and government expenditure have no significant impact on import.

This is in line with the findings of the study that had been obtained by using multiple regression technique, which confirmed the quality of the econometric model.