ATTENTION RECURRENT NEURAL NETWORK WITH EARTHWORM OPTIMIZATION ON GROSS DOMESTIC PRODUCT PREDICTION USING MAIN ECONOMIC ACTIVITIES

Abstract

Gross domestic product (GDP) is a vital metric for evaluating the financial strength and development of a nation. It extends the complete value of services and goods produced within an exact time, presenting critical perceptions into the complete financial performance and health. This study focuses on enhancing GDP prediction by examining key economic activities such as non-oil, oil, and govern- ment sectors. Understanding these modules is important for accurately predicting economic trials, which impact tax revenue, living standards, and economic stabil- ity. By incorporating these foremost financial activities, the research emphasizes improving the exactness of GDP prediction and provides actionable perceptions for strategic economic policy and planning growth. Besides, the study inspects how variations in these areas affect GDP, giving a more complete view of trade trends and helping shareholders make informed decisions to raise steady growth. This study proposes GDP prediction by utilizing an attention recurrent neural network with earthworm optimization algorithm (GDPP-ARNNEOA). The main objective of the GDPP-ARNNEOA technique is to improve GDP prediction accuracy by an- alyzing key economic activities to inform economic planning and policy-making. To accomplish that, the GDPP-ARNNEOA approach performs normalization by utilizing a min-max scaler. Then, the ARNN approach is employed for prediction process. Subsequently, the GDPP-ARNNEOA model accomplishes the hyperpa- rameter tuning by implementing the EOA model. The performance validation of the GDPP-ARNNEOA technique is examined in terms of various measures namely Mean squared error, mean absolute error, and mean absolute percentage error. The experimental results revealed the superior performance of the GDPP-ARNNEOA technique over other recent models.