Controlling Shareholders Management Risk Warning Based on Flexible Neural Tree

This paper studies the early-warning of controlling shareholders management risk of Chinese listed companies, using the optimization algorithm of the improved structure of flexible neural tree model to build early-warning model of risk. In the premise of higher accuracy, it selects the key risk factors and the optimal model. In the learning process of Flexible neural tree model, the evolution generation of Algorithm is not a fixed value, but the error rate to control the evolution generation. Tests prove that the algorithm is the optimal model, the maximum efficiency and accuracy. The structure of Flexible neural tree model and the parameter optimization were completed by the probabilistic incremental program evolution and the simulation annealing, respectively. The results show that the method is a very effective for early-warning controlling shareholder management risk.