在这篇文章中,我们将使用Python和Keras库来构建一个简单的卷积神经网络(CNN),用于识别MNIST数据集中的手写数字。
from keras.datasets import mnist
from keras.models import Sequential
from keras.layers import Dense, Dropout, Flatten
from keras.layers import Conv2D, MaxPooling2D
from keras import backend as K
# 配置学习参数
batch_size = 128
num_classes = 10
epochs = 10
# 输入数据
img_rows, img_cols = 28, 28
# 载入MNIST数据集
(x_train, y_train), (x_test, y_test) = mnist.load_data()
if K.image_data_format() == 'channels_first':
x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols)
x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols)
input_shape = (1, img_rows, img_cols)
else:
x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)
x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)
input_shape = (img_rows, img_cols, 1)
# 将像素值标准化
x_train = x_train.astype('float32') / 255
x_test = x_test.astype('float32') / 255
# 转换类向量为二值类标签
y_train = keras.utils.to_categorical(y_train, num_classes)
y_test = keras.utils.to_categorical(y_test, num_classes)
# 定义CNN模型
model = Sequential()
# 卷积层
model.add(Conv2D(32, kernel_size=(3, 3),
activation='relu',
input_shape=input_shape))
model.add(Conv2D(64, (3, 3), activation='relu'))
# 最大池化层
model.add(MaxPooling2D(pool_size=(2, 2)))
# 随机丢弃层
model.add(Dropout(0.25))
# 平铺层
model.add(Flatten())
# 全连接层
model.add(Dense(128, activation='relu'))
model.add(Dropout(0.5))
# 输出层
model.add(Dense(num_classes, activation='softmax'))
# 编译模型
model.compile(loss=keras.losses.categorical_crossentropy,
optimizer=keras.optimizers.Adadelta(),
metrics=['accuracy'])
# 训练模型
model.fit(x_train, y_train,
batch_size=batch_size,
epochs=epochs,
verbose=1,
validation_data=(x_test, y_test))
# 评估模型
score = model.evaluate(x_test, y_test, verbose=0)
print('Test loss:', score[0])
print('Test accuracy:', score[1])
这段代码展示了如何使用Keras库来构建和训练一个简单的CNN模型,用于MNIST数据集的手写数字识别。代码配置了学习参数,加载了MNIST数据集,对数据进行了预处理,定义了CNN模型的结构,编译并训练了模型,最后评估了模型的性能。这是一个很好的入门级别的CNN示例,适合初学者学习