Keras在一个等级Cifar10上过适合

Question

为了让所有的事情都清楚，让我展示一下整个模型，非常简单：

from keras.datasets import cifar10 #much more libraries imported
# simple prerocessing 
(x_train, y_train), (x_test, y_test) = cifar10.load_data()    
batch_size = 32
num_classes = 10
y_train = np_utils.to_categorical(y_train, num_classes)
y_test = np_utils.to_categorical(y_test, num_classes)
x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train  /= 255
x_test /= 255

def base_model():

    model = Sequential()
    model.add(Conv2D(32, (3, 3), padding='same', input_shape=x_train.shape[1:]))
    model.add(Activation('relu'))
    model.add(Conv2D(32,(3, 3)))
    model.add(Activation('relu'))
    model.add(MaxPooling2D(pool_size=(2, 2)))
    model.add(Dropout(0.25))

    model.add(Conv2D(64, (3, 3), padding='same'))
    model.add(Activation('relu'))
    model.add(Conv2D(64, (3,3)))
    model.add(Activation('relu'))
    model.add(MaxPooling2D(pool_size=(2, 2)))
    model.add(Dropout(0.25))

    model.add(Flatten())
    model.add(Dense(512))
    model.add(Activation('relu'))
    model.add(Dropout(0.5))
    model.add(Dense(num_classes))
    model.add(Activation('softmax'))

    sgd = SGD(lr = 0.1, decay=1e-6, momentum=0.9, nesterov=True)
    # Train model

    model.compile(loss='categorical_crossentropy', optimizer=sgd, metrics=['accuracy'])
    return model

cnn_n = base_model()
cnn_n.summary()

# Fit model

cnn = cnn_n.fit(x_train, y_train, batch_size=batch_size, epochs=epochs, validation_data=(x_test,y_test)
                ,shuffle=True, verbose= 
0)

正如您所看到的，训练错误和验证甚至不能减少错误

sequential_model_to_ascii_printout(cnn_n)
 OPERATION           DATA DIMENSIONS   WEIGHTS(N)   WEIGHTS(%)

               Input   #####     32   32    3
              Conv2D    \|/  -------------------       896     0.1%
                relu   #####     32   32   32
              Conv2D    \|/  -------------------      9248     0.7%
                relu   #####     30   30   32
        MaxPooling2D   Y max -------------------         0     0.0%
                       #####     15   15   32
             Dropout    | || -------------------         0     0.0%
                       #####     15   15   32
              Conv2D    \|/  -------------------     18496     1.5%
                relu   #####     15   15   64
              Conv2D    \|/  -------------------     36928     3.0%
                relu   #####     13   13   64
        MaxPooling2D   Y max -------------------         0     0.0%
                       #####      6    6   64
             Dropout    | || -------------------         0     0.0%
                       #####      6    6   64
             Flatten   ||||| -------------------         0     0.0%
                       #####        2304
               Dense   XXXXX -------------------   1180160    94.3%
                relu   #####         512
             Dropout    | || -------------------         0     0.0%
                       #####         512
               Dense   XXXXX -------------------      5130     0.4%
             softmax   #####          10

混淆矩阵，模型肯定比第三类强：

YU测试还包含其他类：

y_test
array([[0., 0., 0., ..., 0., 0., 0.],
       [0., 0., 0., ..., 0., 1., 0.],
       [0., 0., 0., ..., 0., 1., 0.],
       ...,
       [0., 0., 0., ..., 0., 0., 0.],
       [0., 1., 0., ..., 0., 0., 0.],
       [0., 0., 0., ..., 1., 0., 0.]]

为什么模特“看”只有一节课？你知道吗

附言：我遵循这个指南：https://blog.plon.io/tutorials/cifar-10-classification-using-keras-tutorial/

Answer 1

我觉得这个CIFAR-10任务可以选择Adam优化算法，SGD收敛速度更快。而你设定的学习率太大（你可以设定lr=0.01或lr=0.001），会接近最小学习点震惊。这个是我的代码：CIFAR-10