如何在 TensorFlow 中添加 dropout 层？以及如何在 python 3.x 中增加 Numpy 数组？

Question

首先，我是这个领域的新手，我正在尝试添加 dropout 层以查看模型中性能的变化。我无法弄清楚我应该在下面的代码中在哪里以及如何添加一个 dropout 层。 此外，我想对大小为 39*200 的 numpy 数组进行数据扩充（移位），以使第一列转移到第二列，第二列转移到第三列，依此类推。最后一个转移第一个。这就像剪切图像的最后一部分并将其粘贴到第一部分。

def conv2d(x, W, b, strides=1):
x = tf.nn.conv2d(x, W, strides=[1, strides, strides, 1], padding='SAME')
x = tf.nn.bias_add(x, b)
return tf.nn.relu(x) 

def maxpool2d(x, k=2):
return tf.nn.max_pool(x, ksize=[1, k, k, 1], strides=[1, k, k, 1],padding='SAME')

weights = {
'wc1': tf.get_variable('W0', shape=(3,3,1,32), initializer=tf.contrib.layers.xavier_initializer()), 
'wc2': tf.get_variable('W1', shape=(3,3,32,64), initializer=tf.contrib.layers.xavier_initializer()), 
'wc3': tf.get_variable('W2', shape=(3,3,64,32), initializer=tf.contrib.layers.xavier_initializer()), 
'wc4': tf.get_variable('W3', shape=(3,3,32,128), initializer=tf.contrib.layers.xavier_initializer()),
'wc5': tf.get_variable('W4', shape=(3,3,128,64), initializer=tf.contrib.layers.xavier_initializer()),
'wd1': tf.get_variable('W7', shape=(4*4*56,64), initializer=tf.contrib.layers.xavier_initializer()), 
'out': tf.get_variable('W8', shape=(64,n_classes), initializer=tf.contrib.layers.xavier_initializer()), 

}

biases = {
'bc1': tf.get_variable('B0', shape=(32), initializer=tf.contrib.layers.xavier_initializer()),
'bc2': tf.get_variable('B1', shape=(64), initializer=tf.contrib.layers.xavier_initializer()),
'bc3': tf.get_variable('B2', shape=(32), initializer=tf.contrib.layers.xavier_initializer()),
'bc4': tf.get_variable('B3', shape=(128), initializer=tf.contrib.layers.xavier_initializer()),
'bc5': tf.get_variable('B4', shape=(64), initializer=tf.contrib.layers.xavier_initializer()),
'bd1': tf.get_variable('B7', shape=(64), initializer=tf.contrib.layers.xavier_initializer()),
'out': tf.get_variable('B8', shape=(2), initializer=tf.contrib.layers.xavier_initializer()),

}

def conv_net(x, weights, biases):  
conv1 = conv2d(x, weights['wc1'], biases['bc1'])
conv1 = maxpool2d(conv1, k=2)

conv2 = conv2d(conv1, weights['wc2'], biases['bc2'])
conv2 = maxpool2d(conv2, k=2)

conv3 = conv2d(conv2, weights['wc3'], biases['bc3'])
conv3 = maxpool2d(conv3, k=2)

conv4 = conv2d(conv3, weights['wc4'], biases['bc4'])
conv4 = maxpool2d(conv4, k=2)

conv5 = conv2d(conv4, weights['wc5'], biases['bc5'])
conv5 = maxpool2d(conv5, k=2)

fc1 = tf.reshape(conv5, [-1, weights['wd1'].get_shape().as_list()[0]])
fc1 = tf.add(tf.matmul(fc1, weights['wd1']), biases['bd1'])
fc1 = tf.nn.relu(fc1) 
out = tf.add(tf.matmul(fc1, weights['out']), biases['out'])
return out

pred = conv_net(x, weights, biases)

cost =tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=pred, 
labels=y), name='Cost')



optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(cost)


correct_prediction = tf.equal(tf.argmax(pred, 1), tf.argmax(y, 1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))

init = tf.global_variables_initializer(),


with tf.Session() as sess:
sess.run(init)
train_loss = []
test_loss = []
train_accuracy = []
test_accuracy = []

if not os.path.exists('summaries'):
    os.mkdir('summaries')
if not os.path.exists(os.path.join('summaries','first')):
    os.mkdir(os.path.join('summaries','first'))

summary_writer = tf.summary.FileWriter(os.path.join('summaries','first'), sess.graph)  

for i in range(training_iters):
    for batch in range(len(X_train)//batch_size):
        batch_x = X_train[batch*batch_size:min((batch+1)*batch_size,len(X_train))]
        batch_y = Y_train[batch*batch_size:min((batch+1)*batch_size,len(Y_train))]    
        opt = sess.run(optimizer, feed_dict={x: batch_x,
                                                          y: batch_y})
        loss, acc = sess.run([cost, accuracy], feed_dict={x: batch_x,
                                                          y: batch_y})
    print("Iter " + str(i) + ", Loss= " + \
                  "{:.6f}".format(loss) + ", Training Accuracy= " + \
                  "{:.5f}".format(acc))
    print("Optimization Finished!")

    test_acc,valid_loss = sess.run([accuracy,cost], feed_dict={x: X_test,y : Y_test})
    train_loss.append(loss)
    test_loss.append(valid_loss)
    train_accuracy.append(acc)
    test_accuracy.append(test_acc)

    print("Testing Accuracy:","{:.5f}".format(test_acc))
    print("Accuracy:", accuracy.eval({x: X_test, y: Y_test}))

代码链接：[1]：https://drive.google.com/file/d/1BcbLAlVG0QR8QKToyij9gniQ7E9gvaCc/view?usp=sharing

Answer 1

辍学

你可以将 dropout 层放在 max-pooling 之后，类似这样：

# _____________ FIRST MAX POOLING LAYER _____________________
A_pool1 = tf.nn.max_pool(A_conv1)

# _____________ FIRST DROPOUT LAYER _____________________
A_out1 = tf.nn.dropout(x=A_pool1, rate=dropout_prob)

# _____________ SECOND CONVOLUTIONAL LAYER _____________________
A_conv2 = tf.nn.relu(tf.nn.conv2d(A_out1, W_conv2))

其中 dropout_prob 是 x 的每个元素被丢弃的概率。

另一个例子是here，他们将 dropout 层放在密集层之后（最后）。

在您的具体情况下，您可以这样做：

conv2 = maxpool2d(conv2, k=2)
A_out1 = tf.nn.dropout(x=conv2, rate=0.5)
conv3 = conv2d(A_out1, weights['wc3'], biases['bc3'])

数据增强

要做到这一点，你可以使用 Numpy 函数 roll ，这里解释为 Numpy roll

array([8, 9, 0, 1, 2, 3, 4, 5, 6, 7])
x2 = np.reshape(x, (2,5))

>>> x2
array([[0, 1, 2, 3, 4],
   [5, 6, 7, 8, 9]])

np.roll(x2, 1, axis=1)
array([[4, 0, 1, 2, 3],
       [9, 5, 6, 7, 8]])