Text_cnn

1.基本文本卷积

• For more information refer to:
  Kim 2014
  
  

import numpy as np
import matplotlib.pyplot as plt
from tensorflow import keras
from tensorflow.keras import layers
from tensorflow.keras.preprocessing.sequence import pad_sequences

num_features = 3000
sequence_length = 300
embedding_dimension = 100
(x_train, y_train), (x_test, y_test) = keras.datasets.imdb.load_data(num_words=num_features)

x_train = pad_sequences(x_train, maxlen=sequence_length)
x_test = pad_sequences(x_test, maxlen=sequence_length)
print(x_train.shape)#(25000, 300)
print(x_test.shape)#(25000, 300)
print(y_train.shape)#(25000,)
print(y_test.shape)#(25000,)

2.构造基本句子分类器

def imdb_cnn():
    model=keras.Sequential([
        layers.Embedding(input_dim=num_features,output_dim=embedding_dimension,input_length=sequence_length),
        layers.Conv1D(filters=50,kernel_size=5,strides=1,padding='valid'),#卷积
        layers.MaxPool1D(2,padding='valid'),#池化
        layers.Flatten(),#全连接
        layers.Dense(10,activation='relu'),
        layers.Dense(1,activation='sigmoid')#2分类分类层
                      ])
    #模型配置
    model.compile(optimizer=keras.optimizers.Adam(1e-3),
                 loss=keras.losses.BinaryCrossentropy(),
                 metrics=['accuracy'])
    return model

plt.plot(history.history['accuracy'])
plt.plot(history.history['val_accuracy'])
plt.legend(['training', 'valiation'], loc='upper left')
plt.show()

3.多核卷积网络

filter_sizes=[3,4,5]
def convolution():
    inn = layers.Input(shape=(sequence_length,embedding_dimension,1))#(300,100)
    cnns=[]
    for size in filter_sizes:
        conv = layers.Conv2D(filters=64,kernel_size=(size,embedding_dimension),#([3,4,5],100)
                            strides=1,padding='valid',activation='relu')(inn)
        pool = layers.MaxPool2D(pool_size=(sequence_length-size+1,1),padding='valid')(conv)
        cnns.append(pool)
    outt = layers.concatenate(cnns)
    model = keras.Model(inputs=inn,outputs=outt)
    return model

def cnn_mulfilter():
    model = keras.Sequential([
        layers.Embedding(input_dim = num_features,output_dim=embedding_dimension,
                        input_length=sequence_length),#(300,100)
        layers.Reshape((sequence_length,embedding_dimension,1)),#(300,100,1)
        convolution(),#up
        layers.Flatten(),
        layers.Dense(10,activation='relu'),
        layers.Dropout(0.2),
        layers.Dense(1,activation='sigmoid')
    ])
    model.compile(optimizer=keras.optimizers.Adam(),
                 loss=keras.losses.BinaryCrossentropy(),
                 metrics=['accuracy'])
    return model
model = cnn_mulfilter()
model.summary()