3 33 3 研究方法 RNN 可以分为完全回馈式类神经网路（fully recurrent neural network , FRNN） 图4 及部分回馈式类神经网路（partially recurrent neural network...
基于4个网页-相关网页
Diagonal recurrent neural network (DRNN) is a modified model of the fully connected recurrent neural network with the advantage in capturing the dynamic behavior of a system.
对角循环神经网络是一类经过修正的全连接循环神经网络，在系统动态行为的俘获方面具有明显的优势。
A new training approach for the training algorithm of a fully connected recurrent neural network based on the digital filter theory is proposed.
一种新的基于数字滤波器理论的全互连复值递归神经网络训练方法被提出。
应用推荐