Research on Upstream and Downstream Supply Chain Risk Based on Improved Sparrow Search Algorithm and Long Short-Term Memory Network

This paper presents a method based on improved Long Short Term memory network (LSTM) to predict risks in upstream and downstream supply chains. The key parameters of traditional neural networks are mainly selected by researchers' experience. To solve this problem, Sparrow Search Algorithm (SSA) was introduced to optimize the key parameters and find the optimal model parameters, so as to improve the modeling ability and prediction accuracy of the model on time series data. We verify the effectiveness and superiority of the proposed method by using the actual supply chain data set through empirical research. The results show that our model shows significant performance improvement in identifying and predicting supply chain risks, and provides more accurate decision support for supply chain managers, and improve their competitiveness and market position.