its iter style 輸出 init 向量 數字 ict sha

import tensorflow as tf

# 輸入數據
from tensorflow.examples.tutorials.mnist import input_data

mnist = input_data.read_data_sets("E:\\MNIST_data", one_hot=True)

# 定義網絡的超參數
learning_rate = 0.001
training_iters = 200000
batch_size = 128
display_step = 5

# 定義網絡的參數
# 輸入的維度 (img shape: 28*28)
n_input = 784 
# 標記的維度 (0-9 digits)
n_classes = 10 
# Dropout的概率,輸出的可能性
dropout = 0.75 

# 輸入占位符
x = tf.placeholder(tf.float32, [None, n_input])
y = tf.placeholder(tf.float32, [None, n_classes])
#dropout (keep probability)
keep_prob = tf.placeholder(tf.float32) 

# 定義卷積操作
def conv2d(name,x, W, b, strides=1):
    # Conv2D wrapper, with bias and relu activation
    x = tf.nn.conv2d(x, W, strides=[1, strides, strides, 1], padding=SAME)
    x = tf.nn.bias_add(x, b)
    # 使用relu激活函數
    return tf.nn.relu(x,name=name)  

# 定義池化層操作
def maxpool2d(name,x, k=2):
    # MaxPool2D wrapper
    return tf.nn.max_pool(x, ksize=[1, k, k, 1], strides=[1, k, k, 1],padding=SAME,name=name)

# 規範化操作
def norm(name, l_input, lsize=4):
    return tf.nn.lrn(l_input, lsize, bias=1.0, alpha=0.001 / 9.0,beta=0.75, name=name)

# 定義所有的網絡參數
weights = {
    wc1: tf.Variable(tf.random_normal([11, 11, 1, 96])),
    wc2: tf.Variable(tf.random_normal([5, 5, 96, 256])),
    wc3: tf.Variable(tf.random_normal([3, 3, 256, 384])),
    wc4: tf.Variable(tf.random_normal([3, 3, 384, 384])),
    wc5: tf.Variable(tf.random_normal([3, 3, 384, 256])),
    wd1: tf.Variable(tf.random_normal([4*4*256, 4096])),
    wd2: tf.Variable(tf.random_normal([4096, 1024])),
    out: tf.Variable(tf.random_normal([1024, n_classes]))
}
biases = {
    bc1: tf.Variable(tf.random_normal([96])),
    bc2: tf.Variable(tf.random_normal([256])),
    bc3: tf.Variable(tf.random_normal([384])),
    bc4: tf.Variable(tf.random_normal([384])),
    bc5: tf.Variable(tf.random_normal([256])),
    bd1: tf.Variable(tf.random_normal([4096])),
    bd2: tf.Variable(tf.random_normal([1024])),
    out: tf.Variable(tf.random_normal([n_classes]))
}

# 定義整個網絡
def alex_net(x, weights, biases, dropout):
    # 向量轉為矩陣 Reshape input picture
    x = tf.reshape(x, shape=[-1, 28, 28, 1])

    # 第一層卷積
    # 卷積
    conv1 = conv2d(conv1, x, weights[wc1], biases[bc1])
    # 下采樣
    pool1 = maxpool2d(pool1, conv1, k=2)
    # 規範化
    norm1 = norm(norm1, pool1, lsize=4)

    # 第二層卷積
    # 卷積
    conv2 = conv2d(conv2, norm1, weights[wc2], biases[bc2])
    # 最大池化(向下采樣)
    pool2 = maxpool2d(pool2, conv2, k=2)
    # 規範化
    norm2 = norm(norm2, pool2, lsize=4)

    # 第三層卷積
    # 卷積
    conv3 = conv2d(conv3, norm2, weights[wc3], biases[bc3])
    # 規範化
    norm3 = norm(norm3, conv3, lsize=4)

    # 第四層卷積
    conv4 = conv2d(conv4, norm3, weights[wc4], biases[bc4])

    # 第五層卷積
    conv5 = conv2d(conv5, conv4, weights[wc5], biases[bc5])
    # 最大池化(向下采樣)
    pool5 = maxpool2d(pool5, conv5, k=2)
    # 規範化
    norm5 = norm(norm5, pool5, lsize=4)


    # 全連接層1
    fc1 = tf.reshape(norm5, [-1, weights[wd1].get_shape().as_list()[0]])
    fc1 =tf.add(tf.matmul(fc1, weights[wd1]),biases[bd1])
    fc1 = tf.nn.relu(fc1)
    # dropout
    fc1=tf.nn.dropout(fc1,dropout)

    # 全連接層2
    fc2 = tf.reshape(fc1, [-1, weights[wd2].get_shape().as_list()[0]])
    fc2 =tf.add(tf.matmul(fc2, weights[wd2]),biases[bd2])
    fc2 = tf.nn.relu(fc2)
    # dropout
    fc2=tf.nn.dropout(fc2,dropout)

    # 輸出層
    out = tf.add(tf.matmul(fc2, weights[out]) ,biases[out])
    return out

# 構建模型
pred = alex_net(x, weights, biases, keep_prob)

# 定義損失函數和優化器
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=pred, labels=y))

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

# 評估函數
correct_pred = tf.equal(tf.argmax(pred,1), tf.argmax(y,1))
accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))

# 初始化變量
init = tf.global_variables_initializer()

# 開啟一個訓練
with tf.Session() as sess:
    sess.run(init)
    step = 1
    # 開始訓練,直到達到training_iters,即200000
    while step * batch_size < training_iters:
        #獲取批量數據
        batch_x, batch_y = mnist.train.next_batch(batch_size)
        sess.run(optimizer, feed_dict={x: batch_x, y: batch_y, keep_prob: dropout})
        if step % display_step == 0:
            # 計算損失值和準確度,輸出
            loss,acc = sess.run([cost,accuracy], feed_dict={x: batch_x, y: batch_y, keep_prob: 1.})
            print ("Iter " + str(step*batch_size) + ", Minibatch Loss= " + "{:.6f}".format(loss) + ", Training Accuracy= " + "{:.5f}".format(acc))
        step += 1
    print ("Optimization Finished!")
    # 計算測試集的精確度
    print ("Testing Accuracy:",sess.run(accuracy, feed_dict={x: mnist.test.images[:256],y: mnist.test.labels[:256],keep_prob: 1.}))

技術分享圖片

吳裕雄 python 神經網絡——TensorFlow實現AlexNet模型處理手寫數字識別MNIST數據集