Improvement on function approximation capability of backpropagation neural networks
Improvement on function approximation capability of backpropagation neural networks
复制标题
反向传播神经网络函数逼近能力的改进
DOI：
10.1109/ijcnn.1991.170590
复制
发表时间：
1991
期刊：
影响因子：
--
通讯作者：
Z. Bien
中科院分区：
文献类型：
--
作者：
Jihong Lee;Z. Bien
为了提高隐层单元数给定的网络的逼近精度，作者提出了一种方法，在该方法中，通过误差反向传播来训练激活函数以及权重。他们推广了具有多个参数的S形激活函数，并将参数的一组学习规则导出为误差反向传播的形式。通过一个仿真实例说明了该方法的有效性。<<ETX>>
To increase the approximation accuracy of a network with a given number of units at the hidden layer, the authors propose a method in which the activation functions are trained as well as the weights by error backpropagation. They generalize the sigmoid activation function with several parameters, and derive a set of learning rules for the parameters into the form of error backpropagation. They show the usefulness of the method by a simulation example.<<ETX>>