Prediction of GDP in Time Series Data Based on Exponential Smoothing Model and Neural Network Model

Abstract

Gross Domestic Product (GDP) is the main indicator for measuring the economic size of a country or region, which represents the market value of all final goods and services over a certain period of time. This article compares the performance of exponential smoothing(ES) model and neural network(NN) model in GDP forecasting. The ES model is a classic time series forecasting method widely used in economic data analysis due to its simplicity and ease of use. This study uses single index smoothing and double index smoothing methods to predict GDP, with a focus on analyzing its performance in short-term economic forecasting. To compensate for this deficiency, this study introduced a neural network model to leverage its powerful nonlinear modeling ability and long-term memory capability for time series data. It demonstrates significant advantages in GDP forecasting, better adapting to the complexity of economic data and providing more accurate predictive support for economic decision-making.