tensorflow2.0卷积神经网络MNIST实战

import tensorflow as tf

import pandas as pd

import numpy as np

import matplotlib.pyplot as plt

%matplotlib inline

(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()

img_rows, img_cols = 28, 28

if tf.keras.backend.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)

input_shape = (1, 28, 28)

x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train = x_train / 255
x_test = x_test / 255
x_train.shape

(60000, 28, 28, 1)

# 独热编码
y_train_onehot = tf.keras.utils.to_categorical(y_train)
y_test_onehot = tf.keras.utils.to_categorical(y_test)

# 搭建卷积神经模型
model = tf.keras.Sequential()
model.add(tf.keras.layers.Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(28, 28, 1)))
model.add(tf.keras.layers.MaxPooling2D(pool_size=(2, 2)))
model.add(tf.keras.layers.Conv2D(64, kernel_size=(3, 3), activation='relu'))
model.add(tf.keras.layers.MaxPooling2D(pool_size=(2, 2)))
model.add(tf.keras.layers.Flatten())
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dropout(0.5))
model.add(tf.keras.layers.Dense(10, activation='softmax'))
model.summary()

Model: "sequential_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
conv2d_2 (Conv2D)            (None, 26, 26, 32)        320       
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 13, 13, 32)        0         
_________________________________________________________________
conv2d_3 (Conv2D)            (None, 11, 11, 64)        18496     
_________________________________________________________________
max_pooling2d_3 (MaxPooling2 (None, 5, 5, 64)          0         
_________________________________________________________________
flatten_1 (Flatten)          (None, 1600)              0         
_________________________________________________________________
dense_2 (Dense)              (None, 128)               204928    
_________________________________________________________________
dropout_1 (Dropout)          (None, 128)               0         
_________________________________________________________________
dense_3 (Dense)              (None, 10)                1290      
=================================================================
Total params: 225,034
Trainable params: 225,034
Non-trainable params: 0
_________________________________________________________________

# 设置优化器、损失函数
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])

# 模型训练
history = model.fit(x_train, y_train_onehot, batch_size = 256, epochs = 10, verbose=1, validation_data = (x_test, y_test_onehot))

Epoch 1/10
235/235 [==============================] - 20s 85ms/step - loss: 0.3861 - accuracy: 0.8831 - val_loss: 0.0713 - val_accuracy: 0.9781
Epoch 2/10
235/235 [==============================] - 19s 82ms/step - loss: 0.1077 - accuracy: 0.9687 - val_loss: 0.0496 - val_accuracy: 0.9840
Epoch 3/10
235/235 [==============================] - 21s 89ms/step - loss: 0.0787 - accuracy: 0.9769 - val_loss: 0.0384 - val_accuracy: 0.9874
Epoch 4/10
235/235 [==============================] - 20s 86ms/step - loss: 0.0668 - accuracy: 0.9799 - val_loss: 0.0342 - val_accuracy: 0.9881
Epoch 5/10
235/235 [==============================] - 21s 89ms/step - loss: 0.0547 - accuracy: 0.9832 - val_loss: 0.0331 - val_accuracy: 0.9885
Epoch 6/10
235/235 [==============================] - 19s 81ms/step - loss: 0.0499 - accuracy: 0.9843 - val_loss: 0.0276 - val_accuracy: 0.9903
Epoch 7/10
235/235 [==============================] - 20s 83ms/step - loss: 0.0436 - accuracy: 0.9869 - val_loss: 0.0280 - val_accuracy: 0.9908
Epoch 8/10
235/235 [==============================] - 20s 83ms/step - loss: 0.0404 - accuracy: 0.9879 - val_loss: 0.0265 - val_accuracy: 0.9905
Epoch 9/10
235/235 [==============================] - 20s 83ms/step - loss: 0.0362 - accuracy: 0.9888 - val_loss: 0.0272 - val_accuracy: 0.9912
Epoch 10/10
235/235 [==============================] - 19s 82ms/step - loss: 0.0336 - accuracy: 0.9895 - val_loss: 0.0273 - val_accuracy: 0.9906

score = model.evaluate(x_test, y_test_onehot, verbose=0)
print('Test loss:', score[0])
print('Test accuracy:', score[1])

Test loss: 0.027276480570435524
Test accuracy: 0.9905999898910522