Moldova Republic’s Gross Domestic Product Prevision Using artificial Neural Network Techniques

Considering the present economical status, it is necessary to efficiently manage the Gross Domestic Product (GDP) especially for developing countries. Thus, it will be advantageous to have some knowledge about future trends or specific values of the GDP in accordingly to the dynamics of certain economics indices such as Investment in fixed capital, Imports from other countries etc. The present paper contains the basic research to determine the possibility of using artificial intelligence for the prevision of future GDP. In order to determine the GDP of Moldova Republic the following data were considered: Imports from other countries, Investment in fixed capital, Retail trade and Industrial production. A feed forward artificial neural network (ANN), with 10 hidden neurons, was trained and tested. After 590 iterations a maximum training absolute error of 0.008983 was obtained. Also the absolute validation error was 0.012664 and the network error was 0.000248. The final testing errors belongs to the [0.00; 2.47] interval of absolute values and to the [0.00; 0.23] interval of relative values. The results offer the bases for the future researches.