【机器学习】Tensorflow学习笔记

构建网络模型

基本的MLP网络结构

基本的感知机模型，没有加入b

模型：

Y=W∗X

import tensorflow as tf
import numpy as np
import input_data

# 初始化权重 w
def init_weights(shape):
    return tf.Variable(tf.random_normal(shape, stddev=0.01))

# 定义网络模型，只是基本的mlp模型，堆叠两层的逻辑回归
def model(X, w_h, w_o):
    h = tf.nn.sigmoid(tf.matmul(X, w_h)) 
    return tf.matmul(h, w_o) #这里没有用softmax

# 加载数据
mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)
trX, trY, teX, teY = mnist.train.images, mnist.train.labels, mnist.test.images, mnist.test.labels

# 定义占位符
X = tf.placeholder("float", [None, 784])
Y = tf.placeholder("float", [None, 10])

# 初始化模型参数
w_h = init_weights([784, 625]) 
w_o = init_weights([625, 10])

# 定义模型
py_x = model(X, w_h, w_o)

# 定义损失函数
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(py_x, Y)) 
# 定义训练操作
train_op = tf.train.GradientDescentOptimizer(0.05).minimize(cost) # construct an optimizer
# 定义测试操作
predict_op = tf.argmax(py_x, 1)

# 定义并初始化会话
sess = tf.Session()
init = tf.initialize_all_variables()
sess.run(init)

# 训练测试
for i in range(100):
    for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
        sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end]})
    print i, np.mean(np.argmax(teY, axis=1) ==
                     sess.run(predict_op, feed_dict={X: teX, Y: teY}))

构建多层网络

模型：
多层（3层模型）

import tensorflow as tf
import numpy as np
import input_data

# 初始化权重
def init_weights(shape):
    return tf.Variable(tf.random_normal(shape, stddev=0.01))

# 定义模型，2层的隐藏层+ 3层的dropout
def model(X, w_h, w_h2, w_o, p_drop_input, p_drop_hidden): 
    X = tf.nn.dropout(X, p_drop_input) # 输入就开始用dropout
    h = tf.nn.relu(tf.matmul(X, w_h))

    h = tf.nn.dropout(h, p_drop_hidden) # dropout
    h2 = tf.nn.relu(tf.matmul(h, w_h2))

    h2 = tf.nn.dropout(h2, p_drop_hidden) # dropout

    return tf.matmul(h2, w_o)

# 加载数据
mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)
trX, trY, teX, teY = mnist.train.images, mnist.train.labels, mnist.test.images, mnist.test.labels

# 定义占位符+ 初始化变量
X = tf.placeholder("float", [None, 784])
Y = tf.placeholder("float", [None, 10])

w_h = init_weights([784, 625])
w_h2 = init_weights([625, 625])
w_o = init_weights([625, 10])

# dropout 的概率
p_keep_input = tf.placeholder("float")
p_keep_hidden = tf.placeholder("float")

# 模型
py_x = model(X, w_h, w_h2, w_o, p_keep_input, p_keep_hidden)

# 损失函数
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(py_x, Y))
train_op = tf.train.RMSPropOptimizer(0.001, 0.9).minimize(cost)
predict_op = tf.argmax(py_x, 1)

sess = tf.Session()
init = tf.initialize_all_variables()
sess.run(init)

for i in range(100):
    for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
        sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end],
                                      p_keep_input: 0.8, p_keep_hidden: 0.5})
    print i, np.mean(np.argmax(teY, axis=1) ==
                     sess.run(predict_op, feed_dict={X: teX, Y: teY,
                                                     p_keep_input: 1.0,
                                                     p_keep_hidden: 1.0}))

卷积神经网络

模型：

import tensorflow as tf
import numpy as np
import input_data


def init_weights(shape):
    return tf.Variable(tf.random_normal(shape, stddev=0.01))

# 定义卷积神经网络模型
def model(X, w, w2, w3, w4, w_o, p_keep_conv, p_keep_hidden):
    l1a = tf.nn.relu(tf.nn.conv2d(X, w, [1, 1, 1, 1], 'SAME'))
    l1 = tf.nn.max_pool(l1a, ksize=[1, 2, 2, 1],
                        strides=[1, 2, 2, 1], padding='SAME')
    l1 = tf.nn.dropout(l1, p_keep_conv)

    l2a = tf.nn.relu(tf.nn.conv2d(l1, w2, [1, 1, 1, 1], 'SAME'))
    l2 = tf.nn.max_pool(l2a, ksize=[1, 2, 2, 1],
                        strides=[1, 2, 2, 1], padding='SAME')
    l2 = tf.nn.dropout(l2, p_keep_conv)

    l3a = tf.nn.relu(tf.nn.conv2d(l2, w3, [1, 1, 1, 1], 'SAME'))
    l3 = tf.nn.max_pool(l3a, ksize=[1, 2, 2, 1],
                        strides=[1, 2, 2, 1], padding='SAME')
    l3 = tf.reshape(l3, [-1, w4.get_shape().as_list()[0]])
    l3 = tf.nn.dropout(l3, p_keep_conv)

    l4 = tf.nn.relu(tf.matmul(l3, w4))
    l4 = tf.nn.dropout(l4, p_keep_hidden)

    pyx = tf.matmul(l4, w_o)
    return pyx

# 加载数据
mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)
trX, trY, teX, teY = mnist.train.images, mnist.train.labels, mnist.test.images, mnist.test.labels
trX = trX.reshape(-1, 28, 28, 1)
teX = teX.reshape(-1, 28, 28, 1)

X = tf.placeholder("float", [None, 28, 28, 1])
Y = tf.placeholder("float", [None, 10])

w = init_weights([3, 3, 1, 32])
w2 = init_weights([3, 3, 32, 64])
w3 = init_weights([3, 3, 64, 128])
w4 = init_weights([128 * 4 * 4, 625])
w_o = init_weights([625, 10])

p_keep_conv = tf.placeholder("float")
p_keep_hidden = tf.placeholder("float")
py_x = model(X, w, w2, w3, w4, w_o, p_keep_conv, p_keep_hidden)

# 损失函数
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(py_x, Y))
# 训练操作
train_op = tf.train.RMSPropOptimizer(0.001, 0.9).minimize(cost)
# 测试操作
predict_op = tf.argmax(py_x, 1)

sess = tf.Session()
init = tf.initialize_all_variables()
sess.run(init)

for i in range(100):
    for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
        sess.run(train_op, feed_dict={X: trX[start:end], Y: trY[start:end],
                                      p_keep_conv: 0.8, p_keep_hidden: 0.5})

    test_indices = np.arange(len(teX)) # Get A Test Batch
    np.random.shuffle(test_indices)
    test_indices = test_indices[0:256]

    print i, np.mean(np.argmax(teY[test_indices], axis=1) ==
                     sess.run(predict_op, feed_dict={X: teX[test_indices],
                                                     Y: teY[test_indices],
                                                     p_keep_conv: 1.0,
                                                     p_keep_hidden: 1.0}))