Question

我正在尝试使用CNN架构对文本句子进行分类。网络的体系结构如下：

text_input = Input(shape=X_train_vec.shape[1:], name = "Text_input")

conv2 = Conv1D(filters=128, kernel_size=5, activation='relu')(text_input)
drop21 = Dropout(0.5)(conv2)
pool1 = MaxPooling1D(pool_size=2)(drop21)
conv22 = Conv1D(filters=64, kernel_size=5, activation='relu')(pool1)
drop22 = Dropout(0.5)(conv22)
pool2 = MaxPooling1D(pool_size=2)(drop22)
dense = Dense(16, activation='relu')(pool2)

flat = Flatten()(dense)
dense = Dense(128, activation='relu')(flat)
out = Dense(32, activation='relu')(dense)

outputs = Dense(y_train.shape[1], activation='softmax')(out)

model = Model(inputs=text_input, outputs=outputs)
# compile
model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])

我有一些回调，如early_stopping和reduceLR，以在无效性没有改善（减少）的情况下停止训练并降低学习率。

early_stopping = EarlyStopping(monitor='val_loss', 
                               patience=5)
model_checkpoint = ModelCheckpoint(filepath=checkpoint_filepath,
                                   save_weights_only=False,
                                   monitor='val_loss',
                                   mode="auto",
                                   save_best_only=True)
learning_rate_decay = ReduceLROnPlateau(monitor='val_loss', 
                                        factor=0.1, 
                                        patience=2, 
                                        verbose=1, 
                                        mode='auto',
                                        min_delta=0.0001, 
                                        cooldown=0,
                                        min_lr=0)

模型经过训练后，训练的历史记录如下：

在这里我们可以看到，从第5阶段开始，验证损失并没有改善，并且每个步骤的训练损失都过大。

我想知道我在CNN架构中是否做错了什么？辍学层不足以避免过度拟合吗？减少过拟合的其他方法还有哪些？

有什么建议吗？

谢谢。

编辑：

我也尝试过正则化，结果甚至更糟：

kernel_regularizer=l2(0.01), bias_regularizer=l2(0.01)

编辑2：

我尝试在每次卷积后应用BatchNormalization层，结果是下一个：

norm = BatchNormalization()(conv2)

编辑3：

应用LSTM体系结构后：

text_input = Input(shape=X_train_vec.shape[1:], name = "Text_input")

conv2 = Conv1D(filters=128, kernel_size=5, activation='relu')(text_input)
drop21 = Dropout(0.5)(conv2)
conv22 = Conv1D(filters=64, kernel_size=5, activation='relu')(drop21)
drop22 = Dropout(0.5)(conv22)

lstm1 = Bidirectional(LSTM(128, return_sequences = True))(drop22)
lstm2 = Bidirectional(LSTM(64, return_sequences = True))(lstm1)

flat = Flatten()(lstm2)
dense = Dense(128, activation='relu')(flat)
out = Dense(32, activation='relu')(dense)

outputs = Dense(y_train.shape[1], activation='softmax')(out)

model = Model(inputs=text_input, outputs=outputs)
# compile
model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])

Answer 1

过度拟合可能由多种因素引起，当模型过于适合训练集时就会发生。

要处理此问题，您可以采取以下方法：

添加更多数据

使用数据增强

使用具有良好概括性的架构

添加正则化（主要是辍学，也可以进行L1 / L2正则化）

降低架构复杂性。

为更清楚起见，您可以阅读https://towardsdatascience.com/deep-learning-3-more-on-cnns-handling-overfitting-2bd5d99abe5d

Answer 2

这是在尖叫转移学习。 google-unversal-sentence-encoder非常适合此用例。将模型替换为

import tensorflow_hub as hub 
import tensorflow_text

text_input = Input(shape=X_train_vec.shape[1:], name = "Text_input")

# this next layer might need some tweaking dimension wise, to correctly fit
# X_train in the model
text_input = tf.keras.layers.Lambda(lambda x: tf.squeeze(x))(text_input)
# conv2 = Conv1D(filters=128, kernel_size=5, activation='relu')(text_input)
# drop21 = Dropout(0.5)(conv2)
# pool1 = MaxPooling1D(pool_size=2)(drop21)
# conv22 = Conv1D(filters=64, kernel_size=5, activation='relu')(pool1)
# drop22 = Dropout(0.5)(conv22)
# pool2 = MaxPooling1D(pool_size=2)(drop22)

# 1) you might need `text_input = tf.expand_dims(text_input, axis=0)` here
# 2) If you're classifying English only, you can use the link to the normal `google-universal-sentence-encoder`, not the multilingual one
# 3) both the English and multilingual have a `-large` version. More accurate but slower to train and infer. 
embedded = hub.KerasLayer('https://tfhub.dev/google/universal-sentence-encoder-multilingual/3')(text_input) 

# this layer seems out of place, 
# dense = Dense(16, activation='relu')(embedded) 

# you don't need to flatten after a dense layer (in your case) or a backbone (in my case (google-universal-sentence-encoder))
# flat = Flatten()(dense)

dense = Dense(128, activation='relu')(flat)
out = Dense(32, activation='relu')(dense)

outputs = Dense(y_train.shape[1], activation='softmax')(out)

model = Model(inputs=text_input, outputs=outputs)

Answer 3

我认为，由于您正在执行文本分类，因此添加1或2个LSTM层可能有助于网络更好地学习，因为它可以更好地与数据上下文关联。我建议在展平层之前添加以下代码。

<div class="videobg">
  <video playsinline webkit-playsinline autoplay loop muted>
  <source src="http://techslides.com/demos/sample-videos/small.mp4" type="video/mp4">
</video>
  <img src="https://dummyimage.com/200x200/cc0000/ccc.jpg">
</div>

LSTM层可以帮助神经网络学习某些单词之间的关联，并可以提高网络的准确性。

我还建议删除“最大池”层，因为最大池（尤其是在文本分类中）会导致网络丢弃一些有用的功能。只需保留卷积层和辍学即可。在展平前还要去除密集层并添加上述LSTM。

Answer 4

目前尚不清楚如何将文本输入模型。我假设您标记文本以将其表示为整数序列，但是在将其输入模型之前是否使用任何词嵌入？如果没有，我建议您在模型开始时放置一个可训练的张量流Embedding层。有一种名为“嵌入查找”的聪明技巧可以加快其训练速度，但是您可以将其保存以备后用。尝试将此层添加到模型中。这样，您的Conv1D层就可以更轻松地处理一系列浮点数了。另外，我建议您在每个Conv1D之后扔BatchNormalization，这应该有助于加快收敛和训练速度。

文字分类CNN过拟合训练

4 个答案: