深度学习-手写数字识别代码

Tensorflow2.0

训练模型

import tensorflow as tf
from tensorflow import keras
from tensorflow.keras.datasets import mnist
from tensorflow.keras import layers
(x_train,y_train),(x_test,y_test) = mnist.load_data()
print(x_train.shape,y_train.shape)
# 创建张量
x_train = tf.constant(x_train,dtype=tf.float32)
x_test = tf.constant(x_test,dtype=tf.float32)
# Reshape
x_train4d = tf.reshape(x_train,[x_train.shape[0],28,28,1])
x_test4d = tf.reshape(x_test,[x_test.shape[0],28,28,1])
print(tf.shape(x_train4d),tf.shape(x_test4d))
#图像标准化
x_train4d,x_test4d = x_train4d/255.0,x_test4d/255.0

#模型搭建
model = keras.Sequential([
    layers.Conv2D(filters=16,kernel_size=(5,5),padding='same',input_shape=(28,28,1),activation='relu'),
    layers.MaxPool2D(pool_size=(2,2)),
    layers.Conv2D(filters=36, kernel_size=(5, 5), padding='same', activation='relu'),
    layers.MaxPool2D(pool_size=(2,2)),
    layers.Dropout(0.25),
    layers.Flatten(),
    layers.Dense(128,activation='relu'),
    layers.Dropout(0.5),
    layers.Dense(10,activation='softmax')
])
print(model.summary())

#开始训练
model.compile(loss='sparse_categorical_crossentropy',optimizer='adam', metrics=['accuracy'])
model.fit(x=x_train4d, y=y_train, validation_split=0.2,epochs=20, batch_size=300, verbose=2)
model.save("./model.h5")

模型预测

import tensorflow as tf
from tensorflow import keras
from tensorflow.keras.datasets import mnist
from tensorflow.keras import layers
(x_train,y_train),(x_test,y_test) = mnist.load_data()
print(x_train.shape,y_train.shape)
# 创建张量
x_train = tf.constant(x_train,dtype=tf.float32)
x_test = tf.constant(x_test,dtype=tf.float32)
# Reshape
x_train4d = tf.reshape(x_train,[x_train.shape[0],28,28,1])
x_test4d = tf.reshape(x_test,[x_test.shape[0],28,28,1])
print(tf.shape(x_train4d),tf.shape(x_test4d))
#图像标准化
x_train4d,x_test4d = x_train4d/255.0,x_test4d/255.0

new_model =tf.keras.models.load_model("./model.h5")
print(new_model.summary())
new_model.evaluate(x_train4d,y_test)

预测结果