新闻分类：多分类问题

from keras.datasets import reuters
(train_data,train_labels),(test_data,test_labels) = reuters.load_data(num_words = 10000)
print(len(train_data))
print(len(test_data))

8982
2246

#编码数据
import numpy as np

def vectorize_sequences(sequences,dimension=10000):
    results = np.zeros((len(sequences),dimension))
    for i,sequence in enumerate(sequences):
        results[i,sequence] = 1
    return results

x_train = vectorize_sequences(train_data)
x_test = vectorize_sequences(test_data)

#one_hot编码
import numpy as np

def to_one_hot(labels,dimension=46):
    results = np.zeros((len(labels),dimension)) 
    for i, label in enumerate(labels):
        results[i,label] = 1
    return results

one_hot_train_labels = to_one_hot(train_labels)    #将训练标签向量化
one_hot_test_labels = to_one_hot(test_labels)      #将测试标签向量化

#模型定义
from keras import models
from keras import layers

model = models.Sequential()
model.add(layers.Dense(64,activation='relu',input_shape=(10000,)))
model.add(layers.Dense(64,activation='relu'))
model.add(layers.Dense(46,activation='softmax'))

#编译模型
model.compile(optimizer='rmsprop',
             loss='categorical_crossentropy',
             metrics=['accuracy'])
#留出验证集
x_val= x_train[:1000]
partial_x_train = x_train[1000:]

y_val = one_hot_train_labels[:1000]
partial_y_train = one_hot_train_labels[1000:]

#训练模型
history = model.fit(partial_x_train,
                    partial_y_train,
                    epochs=20,
                    batch_size=512,
                    validation_data=(x_val,y_val))

WARNING:tensorflow:From G:\Anaconda3\lib\site-packages\tensorflow\python\ops\math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use tf.cast instead.
Train on 7982 samples, validate on 1000 samples
Epoch 1/20
7982/7982 [==============================] - 3s 435us/step - loss: 2.6928 - accuracy: 0.5135 - val_loss: 1.7730 - val_accuracy: 0.6320
Epoch 2/20
7982/7982 [==============================] - 2s 240us/step - loss: 1.4467 - accuracy: 0.6978 - val_loss: 1.3406 - val_accuracy: 0.7000
Epoch 3/20
7982/7982 [==============================] - 2s 237us/step - loss: 1.0650 - accuracy: 0.7657 - val_loss: 1.1355 - val_accuracy: 0.7510
Epoch 4/20
7982/7982 [==============================] - 2s 253us/step - loss: 0.8371 - accuracy: 0.8237 - val_loss: 1.0286 - val_accuracy: 0.7960
Epoch 5/20
7982/7982 [==============================] - 2s 234us/step - loss: 0.6661 - accuracy: 0.8628 - val_loss: 0.9638 - val_accuracy: 0.8050
Epoch 6/20
7982/7982 [==============================] - 2s 236us/step - loss: 0.5284 - accuracy: 0.8933 - val_loss: 0.9422 - val_accuracy: 0.8060
Epoch 7/20
7982/7982 [==============================] - 2s 243us/step - loss: 0.4275 - accuracy: 0.9153 - val_loss: 0.9053 - val_accuracy: 0.8120
Epoch 8/20
7982/7982 [==============================] - 2s 235us/step - loss: 0.3472 - accuracy: 0.9287 - val_loss: 0.9271 - val_accuracy: 0.8070
Epoch 9/20
7982/7982 [==============================] - 2s 243us/step - loss: 0.2897 - accuracy: 0.9352 - val_loss: 0.9160 - val_accuracy: 0.8150
Epoch 10/20
7982/7982 [==============================] - 2s 250us/step - loss: 0.2437 - accuracy: 0.9441 - val_loss: 0.9040 - val_accuracy: 0.8170
Epoch 11/20
7982/7982 [==============================] - 2s 238us/step - loss: 0.2095 - accuracy: 0.9495 - val_loss: 0.9372 - val_accuracy: 0.8080
Epoch 12/20
7982/7982 [==============================] - 2s 241us/step - loss: 0.1837 - accuracy: 0.9538 - val_loss: 0.9433 - val_accuracy: 0.8180
Epoch 13/20
7982/7982 [==============================] - 2s 233us/step - loss: 0.1663 - accuracy: 0.9546 - val_loss: 0.9467 - val_accuracy: 0.8140
Epoch 14/20
7982/7982 [==============================] - 2s 250us/step - loss: 0.1518 - accuracy: 0.9518 - val_loss: 0.9894 - val_accuracy: 0.8060
Epoch 15/20
7982/7982 [==============================] - 2s 240us/step - loss: 0.1451 - accuracy: 0.9554 - val_loss: 0.9927 - val_accuracy: 0.8170
Epoch 16/20
7982/7982 [==============================] - 2s 242us/step - loss: 0.1315 - accuracy: 0.9540 - val_loss: 1.0240 - val_accuracy: 0.8050
Epoch 17/20
7982/7982 [==============================] - 2s 244us/step - loss: 0.1249 - accuracy: 0.9592 - val_loss: 1.0458 - val_accuracy: 0.8020
Epoch 18/20
7982/7982 [==============================] - 2s 238us/step - loss: 0.1235 - accuracy: 0.9585 - val_loss: 1.0098 - val_accuracy: 0.8140
Epoch 19/20
7982/7982 [==============================] - 2s 237us/step - loss: 0.1149 - accuracy: 0.9575 - val_loss: 1.0849 - val_accuracy: 0.8050
Epoch 20/20
7982/7982 [==============================] - 2s 235us/step - loss: 0.1110 - accuracy: 0.9574 - val_loss: 1.0951 - val_accuracy: 0.8020

#绘制训练损失和验证损失
import matplotlib.pyplot as plt

#history_dict = history.history

loss = history.history['loss']
val_loss= history.history['val_loss']

epochs = range(1,len(loss_values) + 1)

plt.plot(epochs,loss,'bo',label='Training loss')
plt.plot(epochs,val_loss,'b',label='Validation loss')
plt.title('Training and Validation loss ')
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.legend()    #作用是给图加上图例

plt.show()

#绘制训练精度和验证精度
plt.clf()

accuracy = history.history['accuracy']
val_acc = history.history['val_accuracy']

plt.plot(epochs,accuracy,'bo',label='Training acc')
plt.plot(epochs,val_acc,'b',label='Validation acc')
plt.title('Training and Validation accuracy ')
plt.xlabel('Epochs')
plt.ylabel('Accuracy')
plt.legend()

plt.show()