Weighted instrumental variable method and multi - innovation gradient algorithm overcome the data saturation and bad data.

为克服数据饱和现象和坏数据对参数辨识的影响,引入加权辅助变量法和多新息梯度法.

互联网

Gradient algorithm has been widely used for training the weights of feedforward neural networks.

梯度算法广泛应用于训练前馈神经网络.

互联网

An efficient barrier conjugate gradient algorithm is developed to solve this optimization problem.

同时,本文还提出了一种有效的阻碍共轭梯度算法来求解约束最小点.

互联网