????? 上一講筆者和大家一起學(xué)習(xí)了如何使用 Tensorflow 構(gòu)建一個卷積神經(jīng)網(wǎng)絡(luò)模型。本節(jié)我們將繼續(xù)利用 Tensorflow 的便捷性完成 mnist 手寫數(shù)字?jǐn)?shù)據(jù)集的識別實戰(zhàn)。mnist 數(shù)據(jù)集是 Yann Lecun 大佬基于美國國家標(biāo)準(zhǔn)技術(shù)研究所構(gòu)建的一個研究深度學(xué)習(xí)的手寫數(shù)字的數(shù)據(jù)集歪今。mnist 由 70000 張不同人手寫的 0-9 10個數(shù)字的灰度圖組成。本節(jié)筆者就和大家一起研究如何利用 Tensorflow 搭建一個 CNN 模型來識別這些手寫的數(shù)字腕柜。
數(shù)據(jù)導(dǎo)入
????? mnist 作為標(biāo)準(zhǔn)深度學(xué)習(xí)數(shù)據(jù)集饶氏,在各大深度學(xué)習(xí)開源框架中都默認(rèn)有進行封裝察滑。所以我們直接從 Tensorflow 中導(dǎo)入相關(guān)的模塊即可:
importtensorflowastf
fromtensorflow.examples.tutorials.mnist
importinput_data
# load mnist data
mnist = input_data.read_data_sets('MNIST_data', one_hot=True)
快速搭建起一個簡易神經(jīng)網(wǎng)絡(luò)模型
????? 數(shù)據(jù)導(dǎo)入之后即可按照 Tensorflow 的范式創(chuàng)建相應(yīng)的 Tensor 變量然后創(chuàng)建會話:
# create the session
sess = tf.InteractiveSession()
# create variables and run the session
x = tf.placeholder('float', shape=[None,784])y_ = tf.placeholder('float', shape=[None,10])W = tf.Variable(tf.zeros([784,10]))b = tf.Variable(tf.zeros([10]))sess.run(tf.global_variables_initializer())
????? 定義前向傳播過程和損失函數(shù):
# define the net and loss functiony = tf.nn.softmax(tf.matmul(x, W) + b)cross_entropy = -tf.reduce_sum(y_*tf.log(y))
????? 進行模型訓(xùn)練:
# train the model
train_step = tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy)
foriinrange(1000): ?batch = mnist.train.next_batch(50) ?train_step.run(feed_dict={x: batch[0], y_: batch[1]})
????? 使用訓(xùn)練好的模型對測試集進行預(yù)測:
# evaluate the model
correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_,1))accuracy = tf.reduce_mean(tf.cast(correct_prediction,"float"))print(accuracy.eval(feed_dict={x: mnist.test.images, y_: mnist.test.labels}))
????? 預(yù)測準(zhǔn)確率為 0.9前弯,雖然說也是一個很高的準(zhǔn)確率了蚪缀,但對于 mnist 這種標(biāo)準(zhǔn)數(shù)據(jù)集來說,這樣的結(jié)果還有很大的提升空間恕出。所以我們繼續(xù)優(yōu)化模型結(jié)構(gòu)询枚,為模型添加卷積結(jié)構(gòu)。
搭建卷積神經(jīng)網(wǎng)絡(luò)模型
????? 定義初始化模型權(quán)重函數(shù):
# initilize the weight
defweight_variable(shape): ? ?initial = tf.truncated_normal(shape, stddev=0.1)
returntf.Variable(initial)
defbias_variable(shape): ? ?initial = tf.constant(0.1, shape=shape)
returntf.Variable(initial)
????? 定義卷積和池化函數(shù):
# convolutional and pooling
defconv2d(x, W):
returntf.nn.conv2d(x, W, strides=[1,1,1,1], padding='SAME')
def
max_pool_2x2(x):
returntf.nn.max_pool(x, ksize=[1,2,2,1], ? ? ? ? ? ? ? ? ? ? ? ?strides=[1,2,2,1], padding='SAME')
????? 搭建第一層卷積:
# the first convolution layer
W_conv1 = weight_variable([5,5,1,32])b_conv1 = bias_variable([32])x_image = tf.reshape(x, [-1,28,28,1])h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)h_pool1 = max_pool_2x2(h_conv1)
????? 搭建第二層卷積:
# the second convolution layer
W_conv2 = weight_variable([5,5,32,64])b_conv2 = bias_variable([64])h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)h_pool2 = max_pool_2x2(h_conv2)
????? 搭建全連接層:
# dense layer/full_connected layer
W_fc1 = weight_variable([7*7*64,1024])b_fc1 = bias_variable([1024])h_pool2_flat = tf.reshape(h_pool2, [-1,7*7*64])h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)
????? 設(shè)置 dropout 防止過擬合:
# dropout to prevent overfitting
keep_prob = tf.placeholder("float")h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)
????? 對輸出層定義 softmax :
# model output
W_fc2 = weight_variable([1024,10])b_fc2 = bias_variable([10])y_conv = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2) + b_fc2)
????? 訓(xùn)練模型并進行預(yù)測:
# model trainning and evaluating
cross_entropy = -tf.reduce_sum(y_*tf.log(y_conv))train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)correct_prediction = tf.equal(tf.argmax(y_conv,1), tf.argmax(y_,1))accuracy = tf.reduce_mean(tf.cast(correct_prediction,"float"))sess.run(tf.initialize_all_variables())
foriinrange(20000): ? ?batch = mnist.train.next_batch(50)
ifi%100==0: ? ? ? ? ? ?train_accuracy = accuracy.eval(feed_dict={ ? ? ? ? ? ? ? ?x:batch[0], y_: batch[1], keep_prob:1.0}) ? ?print("step %d, training accuracy %g"%(i, train_accuracy)) ? ?train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob:0.5})print("test accuracy %g"%accuracy.eval(feed_dict={ ? ?x: mnist.test.images, y_: mnist.test.labels, keep_prob:1.0}))
????? 部分迭代過程和預(yù)測結(jié)果如下:
????? 經(jīng)過添加兩層卷積之后我們的模型預(yù)測準(zhǔn)確率達到了 0.9931浙巫,模型訓(xùn)練的算是比較好了金蜀。
? 注:本深度學(xué)習(xí)筆記系作者學(xué)習(xí) Andrew NG 的 deeplearningai 五門課程所記筆記,其中代碼為每門課的課后assignments作業(yè)整理而成的畴。