tensorflow手写数字识别（函数式写法）

手写数字识别（函数式写法）

from tensorflow.keras.datasets import mnist
from tensorflow.keras.layers import Dense, Conv2D, MaxPooling2D, Flatten, Input
from tensorflow.keras.utils import to_categorical
import numpy as np
import matplotlib.pyplot as plt
from tensorflow.python.keras import Model

# 1.准备数据
(train_x, train_y), (test_x, test_y) = mnist.load_data()
train_x = np.expand_dims(train_x, axis=3)
train_x = train_x.astype('f2')/255
test_x = np.expand_dims(test_x, axis=3)
test_x = test_x.astype('f2')/255
train_y = to_categorical(train_y)
test_y = to_categorical(test_y)

input_tensor = Input(shape=(28, 28, 1))
conv2d = Conv2D(filters=32, kernel_size=(3, 3), activation="relu")(input_tensor)
maxPooling2D = MaxPooling2D(pool_size=(2, 2))(conv2d)
conv2d = Conv2D(filters=32, kernel_size=(3, 3), activation="relu")(maxPooling2D)
maxPooling2D = MaxPooling2D(pool_size=(2, 2))(conv2d)
conv2d = Conv2D(filters=32, kernel_size=(3, 3), activation="relu")(maxPooling2D)
flatten = Flatten()(conv2d)
output_tensor = Dense(units=10, activation='softmax')(flatten)
model = Model(input_tensor, output_tensor)
model.compile(optimizer='rmsprop', loss="categorical_crossentropy", metrics=["accuracy"])
history = model.fit(train_x, train_y, batch_size=128, epochs=20, validation_data=(test_x, test_y))

plt.rcParams['font.family'] = ['sans-serif']
plt.rcParams['font.sans-serif'] = ['SimHei']
accuracy = history.history['accuracy']
val_accuracy = history.history['val_accuracy']
epochs = range(1, len(accuracy)+1)
plt.plot(epochs, accuracy, label='训练精度', c = 'r')
plt.plot(epochs, val_accuracy, label='验证精度', c = 'b')
plt.xlabel('epochs')
plt.ylabel('accuracy')
plt.title('author:peiyuanman')
plt.legend()
plt.show()