循环神经网络---卷积、池化、GRU

卷积-池化-卷积-GRU

from keras.models import Sequential
from keras import layers
from keras.optimizers import RMSprop

model = Sequential()
model.add(layers.Conv1D(32,5,activation='relu',input_shape=(None,float_data.shape[-1])))
model.add(layers.MaxPooling1D(3))
model.add(layers.Conv1D(32,5,activation='relu')
model.add(layers.GRU(32,dropout=0.1,recurrent_dropout=0.5)
model.add(layers.Dense(1))

# 一维卷积和一维池化
# 精度不高，仅供和二维卷积对比
from keras.models import Sequential
from keras import layers
from keras.optimizers import RMSprop

model = Sequential()
# Embeding后直接跟RNN和一维卷积
model.add(layers.Embedding(max_features,128,input_length=max_len))
# 二维卷积核一般3*3=9个特征向量
# 一维卷积核一般直接取7或9
model.add(layers.Conv1D(32,7,activation='relu'))
model.add(layers.MaxPooling1D(5))
model.add(layers.Conv1D(32,7,activation='relu'))
# 全局池化层,将三维输出转换为二维输出，后接全连接层
# 也可以用Flatten代替全局池化层
# model.add(layers.Flatten())
model.add(layers.GlobalMaxPooling1D())