莫烦大大keras的Mnist手写识别（5）----自编码

一、步骤：

1. 导入包和读取数据

2. 数据预处理

3. 编码层和解码层的建立 + 构建模型

4. 编译模型

5. 训练模型

6. 测试模型【只用编码层来画图】

二、代码：

1、导入包和读取数据

#导入相关的包
import numpy as np
np.random.seed(1337)  # for reproducibility

from keras.datasets import mnist
from keras.models import Model #采用通用模型
from keras.layers import Dense, Input #只用到全连接层
import matplotlib.pyplot as plt

#读取数据
(X_train, _), (X_test, y_test) = mnist.load_data()

2、数据预处理：将28*28维度的数据拉成一个向量784，原数据X_train的shape为（60000，28，28），转成x_train（60000，784）。

x_train = X_train.astype('float32') / 255. - 0.5       # minmax_normalized

x_test = X_test.astype('float32') / 255. - 0.5         # minmax_normalized

x_train = X_train.reshape((x_train.shape[0], -1))

x_test = X_test.reshape((x_test.shape[0], -1))

print(x_train.shape) #(60000, 784)
print(x_test.shape) #(10000, 784)
print(X_train.shape)  # (60000, 28, 28)

3、编码层和解码层的建立+构建模型

# in order to plot in a 2D figure
encoding_dim = 2

# this is our input placeholder
input_img = Input(shape=(784,))


# encoder layers编码层
encoded = Dense(128, activation='relu')(input_img)
encoded = Dense(64, activation='relu')(encoded)
encoded = Dense(10, activation='relu')(encoded)
encoder_output = Dense(encoding_dim)(encoded)

# decoder layers解码层
decoded = Dense(10, activation='relu')(encoder_output)
decoded = Dense(64, activation='relu')(decoded)
decoded = Dense(128, activation='relu')(decoded)
decoded = Dense(784, activation='tanh')(decoded)

#构建模型
#包括编码层也包括解码层
autoencoder = Model(input = input_img,output = decoded)
#只包括编码层
encoder = Model(input = input_img,output = encoder_output)

4、编译模型

#编译模型
autoencoder.compile(optimizer='adam', loss='mse')

5、训练模型【编码和解码一起训练】

autoencoder.fit(x_train, x_train,
                epochs=20,
                batch_size=256,
                shuffle=True)

6、测试模型并画图显示【仅用编码来预测2维的特征空间】

encoded_imgs = encoder.predict(x_test)
plt.scatter(encoded_imgs[:, 0], encoded_imgs[:, 1], c=y_test) #c表示颜色维度
plt.colorbar()
plt.show()

 

转载于:https://www.cnblogs.com/Lee-yl/p/10133354.html