- RNN(循环神经网络)
 
RNN(循环神经网络)
英文  | 中文  | 
Recurrent  | 循环  | 
Recursive  | 递归  | 
weight  | 权重  | 
bias  | 偏置  | 
activation  | 激活值  | 
Dense Layers  | 全连接层  | 
vanishing gradient  | 梯度消失  | 
exploding gradient  | 梯度爆炸  | 
SGD(Stochastic Gradient Descent)  | 随机梯度下降算法  | 
Adam(A Method for Stochastic Optimization)  | 随机优化算法  | 
compile  | 编译  | 
MSE(mean squared error)  | 均方误差  | 
MAE(mean absoluted error)  | 平均绝对误差  | 
MLP(Multilayer Perceptron)  | 多层感知机(一称人工神经网络,ANN)  | 
LSTM(long short-term memory)  | 长短时记忆  | 
GRU(gate recurrent unit)  | 门限循环单元  | 
classification  | 分类  | 
regression  | 回归  | 
truncate  | 截断  | 
BP(back propagation through time)  | 随时间反向传播  | 
peephole  | 窥视孔  | 
Bidirectional  | 双向  | 
softmax  | 柔性最大值  | 
segment  | 部分  | 
architecture  | 架构  | 
hybrid  | 混合  | 
Temporal  | 短时  | 
Connectionist  | 联结  | 
                










