Prediction of stock market index based on ISSA-BP neural network

Stock market index forecasting is a very tempting topic. Appropriate analysis of such a topic will provide valuable insights for investors, traders and policymakers in the appealing stock markets. In this article, an improved sparrow search algorithm is proposed to optimize the initial weight and threshold of BPNN, and is applied to the forecasting of stock market indices. The contributions of this article are mainly in the following aspects: (1) Use the sine chaos model to initialize the population; (2) In the position update formula of the discoverer, the global optimal solution of the previous generation is added, and at the same time, adaptive weights are added to coordinate the capabilities of local mining and global exploration. (3) Combining the Gaussian mutation operator and the reverse learning strategy, perturbation mutation is carried out at the optimal position to generate new solutions. (4) Use the improved sparrow algorithm (ISSA) to optimize the initial weight and threshold of BP neural network. We evaluate the performance of the proposed model on four datasets, which are SSE, SZSE, SP500 and DJI. Two types of models are used to compared with the proposed mode. One of them is the use of swarm intelligence optimization algorithms to optimize BPNN, including GA-BP, PSO-BP, ACO-BP, GWO-BP, CS-BP and SSA-BP. The others are some deep learning models. The experimental results show that the three strategies proposed in this paper greatly improve the optimization ability of the sparrow search algorithm. The ISSA-BP model achieve great success in the short-term prediction of stock prices, which can help investors predict market trends and find the right time to trade. In addition, for policymakers, the rationality of policies can be assessed through market trends, so as to better promote the development of the stock market.