【機器學習】Tensorflow學習筆記
來自： http://blog.csdn.net//chenriwei2/article/details/50615769
構建網絡模型
基本的MLP網絡結構
基本的感知機模型，沒有加入b
模型：
import tensorflow as tf import numpy as np import input_data初始化權重 w
def init_weights(shape): return tf.Variable(tf.random_normal(shape, stddev=0.01))
定義網絡模型，只是基本的mlp模型，堆疊兩層的邏輯回歸
def model(X, w_h, w_o): h = tf.nn.sigmoid(tf.matmul(X, w_h)) return tf.matmul(h, w_o) #這里沒有用softmax
加載數據
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
定義占位符
X = tf.placeholder("float", [None, 784]) Y = tf.placeholder("float", [None, 10])
初始化模型參數
w_h = init_weights([784, 625]) w_o = init_weights([625, 10])
定義模型
py_x = model(X, w_h, w_o)
定義損失函數
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(py_x, Y))
定義訓練操作
train_op = tf.train.GradientDescentOptimizer(0.05).minimize(cost) # construct an optimizer
定義測試操作
predict_op = tf.argmax(py_x, 1)
定義并初始化會話
sess = tf.Session() init = tf.initialize_all_variables() sess.run(init)
訓練測試
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}))</pre>
構建多層網絡
模型：
多層（3層模型）import tensorflow as tf import numpy as np import input_data初始化權重
def init_weights(shape): return tf.Variable(tf.random_normal(shape, stddev=0.01))
定義模型，2層的隱藏層+ 3層的dropout
def model(X, w_h, w_h2, w_o, p_drop_input, p_drop_hidden): X = tf.nn.dropout(X, p_drop_input) # 輸入就開始用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)加載數據
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
定義占位符+ 初始化變量
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)
損失函數
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}))</pre>
卷積神經網絡
模型：
import tensorflow as tf import numpy as np import input_datadef init_weights(shape): return tf.Variable(tf.random_normal(shape, stddev=0.01))
定義卷積神經網絡模型
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加載數據
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)
損失函數
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(py_x, Y))
訓練操作
train_op = tf.train.RMSPropOptimizer(0.001, 0.9).minimize(cost)
測試操作
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}))</pre> </div>