导入数据集
from keras.datasets import mnist
(train_images, train_labels), (test_images, test_labels) = mnist.load_data()

初始数据维度：
> train_images.shape
(60000, 28, 28)
> len(train_labels)
60000
> train_labels
array([5, 0, 4, ..., 5, 6, 8], dtype=uint8)

数据预处理：
train_images = train_images.reshape((60000, 28 * 28))
train_images = train_images.astype('float32') / 255
train_labels = to_categorical(train_labels)
  
之后：
print(train_images, type(train_images), train_images.shape, train_images.dtype)
print(train_labels, type(train_labels), train_labels.shape, train_labels.dtype)
结果：
[[0. 0. 0. ... 0. 0. 0.]
 [0. 0. 0. ... 0. 0. 0.]
 [0. 0. 0. ... 0. 0. 0.]
 ...
 [0. 0. 0. ... 0. 0. 0.]
 [0. 0. 0. ... 0. 0. 0.]
 [0. 0. 0. ... 0. 0. 0.]] <class 'numpy.ndarray'> (60000, 784) float32
[[0. 0. 0. ... 0. 0. 0.]
 [1. 0. 0. ... 0. 0. 0.]
 [0. 0. 0. ... 0. 0. 0.]
 ...
 [0. 0. 0. ... 0. 0. 0.]
 [0. 0. 0. ... 0. 0. 0.]
 [0. 0. 0. ... 0. 1. 0.]] <class 'numpy.ndarray'> (60000, 10) float32

# 将整数序列编码为二进制矩阵
def vectorize_sequences(sequences, dimension=10000):
 # Create an all-zero matrix of shape (len(sequences), dimension)
 results = np.zeros((len(sequences), dimension))
 for i, sequence in enumerate(sequences):
  results[i, sequence] = 1. # set specific indices of results[i] to 1s
 return results


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

第一种方式：shape为(25000,)
y_train = np.asarray(train_labels).astype('float32') #就用这种方式就行了
y_test = np.asarray(test_labels).astype('float32')
第二种方式：shape为(25000,1)
y_train = np.asarray(train_labels).astype('float32').reshape(25000, 1)
y_test = np.asarray(test_labels).astype('float32').reshape(25000, 1)
第三种方式：shape为(25000,2)
y_train = to_categorical(train_labels) #变成one-hot向量
y_test = to_categorical(test_labels)

from keras.datasets import mnist,cifar10，cifar100
(X_train, y_train), (X_valid, Y_valid) = cifar10.load_data() 
model.fit(X_train, Y_train, batch_size=batch_size, nb_epoch=nb_epoch, shuffle=True,
    verbose=1, validation_data=(X_valid, Y_valid), )

from keras.preprocessing.image import ImageDataGenerator
(trainX, trainY), (testX, testY) = cifar100.load_data()
trainX = trainX.astype('float32')
testX = testX.astype('float32')
trainX /= 255.
testX /= 255.
Y_train = np_utils.to_categorical(trainY, nb_classes)
Y_test = np_utils.to_categorical(testY, nb_classes)
generator = ImageDataGenerator(rotation_range=15,
        width_shift_range=5./32,
        height_shift_range=5./32)
generator.fit(trainX, seed=0)
model.fit_generator(generator.flow(trainX, Y_train, batch_size=batch_size),
     steps_per_epoch=len(trainX) // batch_size, epochs=nb_epoch,
     callbacks=callbacks,
     validation_data=(testX, Y_test),
     validation_steps=testX.shape[0] // batch_size, verbose=1)