In recent years, my country’s economic development has entered a new stage, and economic development has entered a new height. Gross Domestic Product (GDP) is the most important and commonly used measure to evaluate social and economic development. It can comprehensively reflect the operating conditions of the economy. At the 2014 APEC summit, Xi Jinping focused on the characteristics of the new normal of China’s economy. The economic growth has undergone major changes, and the economic structure has become more and more complete and has been continuously optimized. Therefore, the driving force of innovation plays an important role in the future social and economic development. The high-tech demonstration zone is the main hub of my country’s economic innovation and scientific and technological progress, and the development of the high-tech demonstration zone is particularly important in the development of the innovative economy. By predicting the economic growth, the development structure can be adjusted in time, and the coordinated allocation of resources can be achieved, which is conducive to the stable and rapid development of the economy. Therefore, it is very necessary to predict my country’s economic growth. At present, more and more scholars have invested more energy in the establishment of economic forecasting models. Previously, the prediction effect of traditional models was not satisfactory. At present, the field of target camera machine learning has been transferred. The application of the powerful nonlinear fitting ability of the machine learning model can improve the prediction effect. Based on this, this paper studies the use of artificial neural network with improved BP algorithm to predict social and economic development, and analyzes social and economic development trends and changing laws. The results show that the forecasting model constructed in this paper has a good forecasting effect, and the research results can provide a theoretical basis for scientifically formulating macro-control policies.