The Macroeconomic Prediction Model based on LSTM Improved Algorithm

In order to solve the overfitting problem caused by large sample size and small sample size of market indicators in macroeconomic forecasting, the LSTM algorithm is adopted to improve the efficiency of economic forecasting. At the same time, the prediction results of the unbiased grey model and the improved grey model are weighted and averaged to be taken into the improved LSTM algorithm model. Through the comparison of prediction effects of various models, it is found that under LSTM model, the difference between the real values and the predicted values of CPI prediction results is 0.3. The introduction of the grey model into the model can make a better scientific prediction of macro economy. And the improved LSTM model has shown better accuracy in predicting GDP. The prediction results of GDP and CPI show that the improved LSTM model can be applied to macroeconomic forecasting.