In order to evaluate the development status of cities and their impact on economic growth, combined with the five new development concepts of innovation, coordination, green, openness and sharing, the evaluation index system of urban high-quality development was constructed, the indicators were screened by OLS least squares method, the impact of high-quality urban development on economic growth and the prediction effect the explored by deep learning method, and the prediction results and accuracy of the multivariate LSTM neural network model and the improved multivariate LSTM neural network model were compared. The results show that the multivariate LSTM neural network model with attention mechanism has better prediction effect, the root mean square error RMSE of the model is reduced by 25%, and the coefficient of determination R² reaches 0.9856.

high-quality urban development  /  deep learning  /  LSTM model  /  attention mechanisms  /  economic growth