Keras学习（三）-Cifar10数据处理（卷积神经网络形式）

1.导入运行库

from keras.datasets import cifar10
import numpy as np
np.random.seed(10)

2.数据预处理

#读入数据
(x_img_train,y_label_train),(x_img_test,y_label_test)=cifar10.load_data()

#显示训练集测试集的形状
print("train data:",'images:',x_img_train.shape,
      " labels:",y_label_train.shape) 
print("test  data:",'images:',x_img_test.shape ,
      " labels:",y_label_test.shape) 

#进行数据标准化处理，提高准确度
x_img_train_normalize = x_img_train.astype('float32') / 255.0
x_img_test_normalize = x_img_test.astype('float32') / 255.0

#对标签值进行one-hot编码
from keras.utils import np_utils
y_label_train_OneHot = np_utils.to_categorical(y_label_train)
y_label_test_OneHot = np_utils.to_categorical(y_label_test)

3.建立训练模型

#导入运行库
from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten
from keras.layers import Conv2D, MaxPooling2D, ZeroPadding2D

#建立堆叠基本模型
model = Sequential()

#卷积1
model.add(Conv2D(filters=32,kernel_size=(3,3),
                 input_shape=(32, 32,3), 
                 activation='relu', 
                 padding='same'))

#dropout层1
model.add(Dropout(rate=0.25))

#池化层1
model.add(MaxPooling2D(pool_size=(2, 2)))

#卷积层2，dropout层2，池化层2
model.add(Conv2D(filters=64, kernel_size=(3, 3), 
                 activation='relu', padding='same'))
model.add(Dropout(0.25))
model.add(MaxPooling2D(pool_size=(2, 2)))

#建立平坦层，全连接隐藏层，输出层
model.add(Flatten())
model.add(Dropout(rate=0.25))

model.add(Dense(1024, activation='relu'))
model.add(Dropout(rate=0.25))

model.add(Dense(10, activation='softmax'))

print(model.summary())#查看训练模型概要

4.如果有已经训练好的模型超参数便读入

try:
    model.load_weights("SaveModel/cifarCnnModelnew1.h5")
    print("加载模型成功!继续训练模型")
except :    
    print("加载模型失败!开始训练一个新模型")

5.训练

#设置训练模型
model.compile(loss='categorical_crossentropy',
              optimizer='adam', metrics=['accuracy'])

#输入训练数据到模型中，进行训练
train_history=model.fit(x_img_train_normalize, y_label_train_OneHot,
                        validation_split=0.2,
                        epochs=10, batch_size=128, verbose=1)   

• veose数值变化对显示的影响
  • 1（动态显示训练过程 ）
    
  • 2（等待完成一个epoch之后才会显示此次epoch的具体信息）
    
• 显示训练过程中的准确率，损失值的变化

import matplotlib.pyplot as plt
def show_train_history(train_acc,test_acc):
    plt.plot(train_history.history[train_acc])
    plt.plot(train_history.history[test_acc])
    plt.title('Train History')
    plt.ylabel('Accuracy')
    plt.xlabel('Epoch')
    plt.legend(['train', 'test'], loc='upper left')
    plt.show()
show_train_history('acc','val_acc')
show_train_history('loss','val_loss')

• 进行模型准确率评估

scores = model.evaluate(x_img_test_normalize, 
                        y_label_test_OneHot, verbose=0)
print(scores[1])

6.进行预测，查看预测结果

label_dict={0:"airplane",1:"automobile",2:"bird",3:"cat",4:"deer",
            5:"dog",6:"frog",7:"horse",8:"ship",9:"truck"}

import matplotlib.pyplot as plt
def plot_images_labels_prediction(images,labels,prediction,
                                  idx,num=10):
    fig = plt.gcf()
    fig.set_size_inches(12, 14)
    if num>25: num=25 
    for i in range(0, num):
        ax=plt.subplot(5,5, 1+i)
        ax.imshow(images[idx],cmap='binary')
                
        title=str(i)+','+label_dict[labels[i][0]]
        if len(prediction)>0:
            title+='=>'+label_dict[prediction[i]]
            
        ax.set_title(title,fontsize=10) 
        ax.set_xticks([]);ax.set_yticks([])        
        idx+=1 
    plt.show()

#显示前10个的预测对比结果
plot_images_labels_prediction(x_img_test,y_label_test,
                              prediction,0,10)

7.查看数据样本预测为各个分类的具体概率

Predicted_Probability=model.predict(x_img_test_normalize)

def show_Predicted_Probability(y,prediction,
                               x_img,Predicted_Probability,i):
    print('label:',label_dict[y[i][0]],
          'predict:',label_dict[prediction[i]])
    plt.figure(figsize=(2,2))
    plt.imshow(np.reshape(x_img_test[i],(32, 32,3)))
    plt.show()
    for j in range(10):
        print(label_dict[j]+
              ' Probability:%1.9f'%(Predicted_Probability[i][j]))

#输出第一个样本对于各个分类的预测概率
show_Predicted_Probability(y_label_test,prediction,
                           x_img_test,Predicted_Probability,0)

8.显示混淆矩阵

y_label_test.reshape(-1)#输入必须是一维数组，转化为一维数组
import pandas as pd
print(label_dict)
pd.crosstab(y_label_test.reshape(-1),prediction,
            rownames=['label'],colnames=['predict'])

9.模型保存（important）参考

• 保存为json

model_json = model.to_json()
with open("SaveModel/cifarCnnModelnew.json", "w") as json_file:
    json_file.write(model_json)

• 保存为h5格式

model.save_weights("SaveModel/cifarCnnModelnew.h5")
print("Saved model to disk")

• 保存为yaml格式

model_yaml = model.to_yaml()
with open("SaveModel/cifarCnnModelnew.yaml", "w") as yaml_file:
    yaml_file.write(model_yaml)