摘要
针对单纯使用神经网络预测物流量出现预测精度低、泛化能力差、耗费时间长等问题,提出一种可以最优化目标的算法,为避免陷入局部最小值,引入遗传算法对神经网络所需要的权值、阈值等可变参数进行最优化处理。仿真结果表明利用遗传算法对神经网络进行优化处理,预测结果能够表现出更高的预测精度与泛化能力。
In view of the problems of inadequate prediction, poor generalizability and time-consumption arising from using neural network alone, the study puts forward an algorithm which can be optimized in order to avoid falling into local minimum, and introduce necessary parameters such as weights and thresholds of the genetic algorithm neural network for optimizing processing. The simulation results show that the neural network using genetic algorithms to optimize the predicted results can exhibit higher prediction accuracy and generalization ability.
出处
《陕西理工学院学报（自然科学版）》
2016年第6期80-85,共6页
Journal of Shananxi University of Technology:Natural Science Edition
关键词
遗传算法
神经网络
物流
城市圈
genetic algorithm
neural network
logistics
city circle
