Stock Index Prediction Based on the PSOPI-BP Neural Network

摘要

In order to improve the prediction ability of the Neuron Network in stock index prediction, we proposed an improve particle swarm neural network algorithm. The Particle Swarm Optimization algorithm based on Parasitic Immune (PSOPI) was used to optimize combination weights of BP neural network model parameters. The BP algorithm was also to obtain the parameters of network further accurate. Finally, experimental results demonstrate the efficacy of our improved algorithm.