其次，从误差反传算法在预测中存在的问题入手，提出一种混合训练算法。
Secondly, the hybrid training algorithm is proposed on the base of the Error Back-propagation algorithm's disadvantage analysis in the paper.
针对多层前馈网络的误差反传算法存在的收敛速度慢，且易陷入局部极小的缺点，提出了采用微粒群算法（PSO）训练多层前馈网络权值的方法。
The particle swarm optimization(PSO) algorithm, is used to train neural network to solve the drawbacks of BP algorithms which is local minimum and slow convergence.
结合小波分析和神经网络的特点，建立了应用于船舶横摇运动时间序列预报的误差反传小波神经网络结构并给出了算法。
Combining wavelet analysis and neural network characteristics, the error back propagation wavelet neural network based structure and algorithm to ship roll time series prediction are given.
经模拟计算，它比传统的基于最陡下降方法的误差反传（SDBEP）算法具有更好的收敛性能。
Simulation computations show that it converges faster than the conventional steepest descent backwards error propagation (SDBEP) algorithm.
获得的值与对应的期望值对比，误差使用反传学习算法传回网络，用以更新连接权值。
Compared obtained value with corresponding expected value, error was sent to network with reverse learning algorithm, so that renovate connect weighting value.
获得的值与对应的期望值对比，误差使用反传学习算法传回网络，用以更新连接权值。
Compared obtained value with corresponding expected value, error was sent to network with reverse learning algorithm, so that renovate connect weighting value.
应用推荐