This paper collects and organizes economic data and establishes ARIMA (Auto Regressive Integrated Moving Average) model and LSTM (Long Short-Term Memory) model to compare their accuracy and stability in predicting the future trend of economic indicators. The results of the study indicate that in some cases, the LSTM model more accurately captures the long-term dependencies in the time series data and improves the forecasting results, while in some cases, the ARIMA model performs more stably and reliably. The study concludes that the LSTM model predicts the results of the indicators of economic management with an accuracy of up to 96%, but it also has shortcomings such as higher complexity. Therefore, choosing the appropriate model depends on the specific data characteristics and forecasting needs, and it is recommended to consider the advantages and disadvantages of ARIMA model and LSTM model in practical applications, and choose the most suitable model for the forecasting work in the field of economic management. This paper provides a more scientific forecasting method and decision-making reference for economic management decision-making.