tensorflow勉強会④ MNIST(難しい方)
mint_deep.pyを動かす
やはりよくわからないので、やはりこの上なくバカっぽく書き直していきます。
新しく出てくる概念
①畳み込み
②プーリング
このページがわかりやすいかも!!!動くし。
deepage.net
③ドロップアウト
密結合層のノードを、ある確率で(普通は50%)使わないようにして学習していきます。
理屈はもろもろありますが、結果的に過学習しづらくなります。
この上なくバカっぽく書き直したコード
import tensorflow as tf import numpy as np # 0.よく使う記述は関数に def conv2d(x, W): # 畳み込み """conv2d returns a 2d convolution layer with full stride.""" return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME') def max_pool_2x2(x): # プーリング """max_pool_2x2 downsamples a feature map by 2X.""" return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME') def weight_variable(shape): """weight_variable generates a weight variable of a given shape.""" initial = tf.truncated_normal(shape, stddev=0.1) return tf.Variable(initial) def bias_variable(shape): """bias_variable generates a bias variable of a given shape.""" initial = tf.constant(0.1, shape=shape) return tf.Variable(initial) # 1.モデルを作成 x = tf.placeholder(tf.float32, [None, 784]) # Reshape to use within a convolutional neural net. # Last dimension is for "features" - there is only one here, since images are # grayscale -- it would be 3 for an RGB image, 4 for RGBA, etc. x_image = tf.reshape(x, [-1, 28, 28, 1]) # First convolutional layer - maps one grayscale image to 32 feature maps. W_conv1 = weight_variable([5, 5, 1, 32]) b_conv1 = bias_variable([32]) h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1) # Pooling layer - downsamples by 2X. h_pool1 = max_pool_2x2(h_conv1) # Second convolutional layer -- maps 32 feature maps to 64. 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) # Second pooling layer. h_pool2 = max_pool_2x2(h_conv2) # Fully connected layer 1 -- after 2 round of downsampling, our 28x28 image # is down to 7x7x64 feature maps -- maps this to 1024 features. 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) # Dropout - controls the complexity of the model, prevents co-adaptation of # features. ドロップアウト keep_prob = tf.placeholder(tf.float32) h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob) # Map the 1024 features to 10 classes, one for each digit W_fc2 = weight_variable([1024, 10]) b_fc2 = bias_variable([10]) y_conv = tf.matmul(h_fc1_drop, W_fc2) + b_fc2 y_ = tf.placeholder(tf.float32, [None, 10]) # 2.学習方法を指定 cross_entropy = tf.reduce_mean( tf.nn.softmax_cross_entropy_with_logits(labels=y_, logits=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, tf.float32)) # 3.コンパイル sess = tf.InteractiveSession() tf.global_variables_initializer().run() # 4.データをロード train_data = np.loadtxt('train-images.txt') train_label_pre = np.loadtxt('train-labels.txt') train_label = np.zeros([60000, 10], dtype=np.float32) for i in range(60000): train_label[i, int(train_label_pre[i])] = 1 test_data = np.loadtxt('test-images.txt') test_label_pre = np.loadtxt('test-labels.txt') test_label = np.zeros([10000, 10], dtype=np.float32) for i in range(10000): test_label[i, int(test_label_pre[i])] = 1 # 5.学習 & 6.学習結果のテスト for iter in range(100): print("iter : " + str(iter)) for i in range( int(60000/100) ): batch_xs = train_data[i*100:(i+1)*100, ] batch_ys = train_label[i*100:(i+1)*100, ] sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys, keep_prob:0.5}) print("test accuracy : " + str(sess.run(accuracy, feed_dict={x: test_data, y_: test_label, keep_prob:1.0})))