Review of swarm intelligence algorithm optimization of BP neural network

As a model used for parameter estimation, BP neural network has remarkable effect in many prediction algorithms. However, there are situations that will fall into the local optimal solution and learning speed is slow. To solve these two problems, this paper combines dynamic adaptive strategy genetic algorithm, method of eliminating honey sources bee colony greedy ant introduction variation operator particle swarm algorithm to improve defects network, It made optimization research on use prediction, also laid groundwork future network.