In this paper, we extract the core idea of state perturbation from
neural networks and define state perturbation formulas to describe the general way of optimization methods. Departing from the core idea and the formulas, we propose a novel optimization method related to neural network structure, named structure perturbation optimization. Our method can produce a structure transforming process to retrain
neural networks to get better problem-solving ability. Experiments validate that our method effectively helps
neural networks to escape from local minima and get superior solutions.