Long Short-Term Memory Neural Intelligent Network Forecast Model Based on Improved Particle Swarm Optimization Algorithm

To improve the accuracy of time series forecasting and enhance the interpretability of the structural feature parameters of the forecasting model, this paper improved Particle Swarm Optimization-Long Short-Term Memory Network model which has been used in ship motion attitude prediction, intelligent traffic flow prediction, stock prediction, etc. Adaptive Particle Swarm Optimization-Long Short-Term Memory neural network model (APSO-LSTM model) is proposed. The prediction model changes the traditional PSO algorithm shortcoming which is easily falling into the local optimal. By constructing an adaptive PSO algorithm to search overall optimal to optimize the key parameters of the LSTM model, the accuracy of the prediction model is improved. Compared with the traditional LSTM model, the impact of human intervention is reduced. This article selects the US gold futures price information as the representative of the gold futures price level to verify the APSO-LSTM prediction model and compares the prediction results of the model with the traditional RNN model and LSTM model prediction results. The results show that the LSTM prediction model based on adaptive PSO improves the accuracy of prediction.