【摘要】 在遗传算法(GeneticAlgorithm)与BP(BackPropagation)网络结构模型相结合的基础上,设计了用遗传算法训练神经网络权重的新方法,并把这种方法用于运动员百米跑成绩预测。与BP算法和LM(LevenbergMarquardt)算法相比,基于混合遗传算法的神经网络不仅有较快的学习速度和较好的学习精度,而且网络的泛化能力(GeneralizationAbility)得到了很大提高。更多还原

【Abstract】 Based upon the combination of Genetic Algorithm and BP Neural Network, this paper puts forward a new approach to learn the weights of BP using GA, and applies this method firstly to predict 100m sprint performance of athletes. Compared with BP algorithm and LM algorithm, the neural network based on hybrid GA not only has quick convergence speed and better learning precision, but also the generalization ability of network has improved greatly.更多还原