循环神经网络TextRNN实现情感短文本分类任务

import numpy as np
from keras.layers import Dense, LSTM, Input, Embedding, Bidirectional
from keras.models import Model
class TextRNN:
    def __init__(self, max_len, num_classes, vocab_size, embedding_dim=128, hidden_dim=64):
        self.max_len = max_len
        self.num_classes = num_classes
        self.vocab_size = vocab_size
        self.embedding_dim = embedding_dim
        self.hidden_dim = hidden_dim
    def build_model(self):
        inputs = Input(shape=(self.max_len,), dtype='int32')
        embed = Embedding(input_dim=self.vocab_size, output_dim=self.embedding_dim, input_length=self.max_len)(inputs)
        rnn = Bidirectional(LSTM(units=self.hidden_dim, return_sequences=True))(embed)
        rnn = Bidirectional(LSTM(units=self.hidden_dim))(rnn)
        outputs = Dense(self.num_classes, activation='softmax')(rnn)
        model = Model(inputs=inputs, outputs=outputs)
        model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
        return model
    def train(self, X_train, y_train, X_test, y_test,batch_size = 64, epochs = 10):
        model = self.build_model()
        model.fit(X_train, y_train, batch_size=batch_size, epochs=epochs, validation_data=(X_test, y_test))
        return model
# 数据预处理
max_len = 100 # 最大文本长度
vocab_size = 5000 # 词汇表大小
embedding_dim = 128 # 嵌入维度
hidden_dim = 64 # RNN隐藏层维度
# 加载数据并进行预处理
from data_loader import DataLoader
data_loader = DataLoader(max_len,max_len,training=False)
X_train, y_train, X_test, y_test, word_index= data_loader.load_data()
vocab_size=len(word_index)
# 进行模型训练
text_rnn = TextRNN(max_len=max_len, num_classes=3, vocab_size=vocab_size,
                       embedding_dim=embedding_dim, hidden_dim=hidden_dim)
model = text_rnn.train(X_train, y_train, X_test, y_test,batch_size=64,epochs=5)
# 进行预测
y_pred = model.predict(X_test)
y_pred = np.argmax(y_pred, axis=1)