深度学习-手写数字识别（卷积神经网络）
概述
代码语言：javascript复制* 数据来源手写数据（kersa）
* 方法：卷积神经网络
* 评估准确率代码
代码语言：javascript复制# 构建卷积层
from keras import layers
from keras import models
model = models.Sequential()
model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)))
# 这里的Conv2D对输入数据进行卷积，输出的为3D的张量
model.add(layers.MaxPooling2D((2, 2)))
# MaxPooling2D与上述类似
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(64, (3, 3), activation='relu'))Using TensorFlow backend.# 添加常规的密集神经网络部分
model.add(layers.Flatten())
model.add(layers.Dense(64, activation='relu'))
model.add(layers.Dense(10, activation='softmax'))
# 通过summary来查看构建的卷积神经网络
model.summary()Model: "sequential_1"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv2d_1 (Conv2D) (None, 26, 26, 32) 320
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 13, 13, 32) 0
_________________________________________________________________
conv2d_2 (Conv2D) (None, 11, 11, 64) 18496
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 5, 5, 64) 0
_________________________________________________________________
conv2d_3 (Conv2D) (None, 3, 3, 64) 36928
_________________________________________________________________
flatten_1 (Flatten) (None, 576) 0
_________________________________________________________________
dense_1 (Dense) (None, 64) 36928
_________________________________________________________________
dense_2 (Dense) (None, 10) 650
=================================================================
Total params: 93,322
Trainable params: 93,322
Non-trainable params: 0
_________________________________________________________________# 带入数据进行运算
# 这部分内容与之前的文章类似
from keras.datasets import mnist
from keras.utils import to_categorical
(train_images, train_labels), (test_images, test_labels) = mnist.load_data()
train_images = train_images.reshape((60000, 28, 28, 1))
train_images = train_images.astype('float32') / 255
test_images = test_images.reshape((10000, 28, 28, 1))
test_images = test_images.astype('float32') / 255
train_labels = to_categorical(train_labels)
test_labels = to_categorical(test_labels)
model.compile(optimizer='rmsprop',
loss='categorical_crossentropy',
metrics=['accuracy'])
model.fit(train_images, train_labels, epochs=5, batch_size=64)Epoch 1/5
60000/60000 [==============================] - 7s 111us/step - loss: 0.1603 - accuracy: 0.9502
Epoch 2/5
60000/60000 [==============================] - 6s 106us/step - loss: 0.0447 - accuracy: 0.9857
Epoch 3/5
60000/60000 [==============================] - 6s 103us/step - loss: 0.0308 - accuracy: 0.9900
Epoch 4/5
60000/60000 [==============================] - 6s 105us/step - loss: 0.0240 - accuracy: 0.9926
Epoch 5/5
60000/60000 [==============================] - 6s 103us/step - loss: 0.0190 - accuracy: 0.9940
<keras.callbacks.callbacks.History at 0x7f951c0e1390># 在测试集评估
test_loss, test_acc = model.evaluate(test_images, test_labels)
test_acc10000/10000 [==============================] - 0s 48us/step
0.9894000291824341在之前的分类学习中，使用普通的神经网络能够达到97.8的精确度，使用卷积神经网络能够达到0.99的精确度
原理解释
- Conv2D 卷积网络和神经网络的区别就在于卷积的过程，第一个卷积层接收一个大小为(28, 28, 1) 的特征图，通过计算32（3*3）个过滤器，输出(26, 26, 32) 的特征图
- MaxPooling2D 最大池化运算就是对特征图进行下采样，使用2×2的窗口和步幅2，卷积使用的是3x3的窗口和步幅1，目的是减少需要处理的特征图的元素个数目
结束语
love&peace
