Forecasting Economic Growth Using a Combination of Wavelet Transform and Perceptron Artificial Neural Network and GMDH (Case Study: Iran ۱۹۶۱-۲۰۲۲)

Economic growth is amongst the most significant macroeconomic indicators, and its forecast form a core component of economic decisions in the short and long-run and at the macro and micro levels of the economy. Also, forecasting economic growth impacts business processes; as economic growth occurs, systematic risk decreases and investment increases. Forecasts economic growth in developing countries is challenging and plays a significant role in planning and policymaking. Thus, predicting economic growth in these countries is of high importance. Iran is considered to be one of the developing countries which has rather unstable tendencies in its economic growth during the last decades and years. This study, using a combination of Wavelet Transform, Perceptron Artificial Neural Network, and Group Method of Data Handling (GMDH), examines Iran's economic growth from ۱۹۶۱ to ۲۰۲۲. In this study, the combination of wavelet transform and artificial neural network perceptron is introduced as an improved model for more accurately predicting Iran's economic growth. The results indicate that this hybrid model, by denoising the input data using wavelets, performs better than previous models and reduces prediction error.

Keywords:

GDP , Forecasting , Wavelet , Perceptron Artificial Neural Network , GMDH