香港科技大学TensorFlow三天速成课件2.pptVIP

TensorFlow @UST Day 2: CNN Sung Kim hunkim+tfust@ Gifts from Google and Line TensorFlow Mechanics feed data and run graph (operation)sess.run (op, feed_dict={x: x_data}) update variables in the graph (and return values) Build graph using TensorFlow operations Machine Learning Basics Linear Regression Logistic Regression (Binary classification) Softmax Classification Neural Networks Linear Regression W = tf.Variable(tf.random_normal([1]), name=weight) b = tf.Variable(tf.random_normal([1]), name=bias) # Our hypothesis XW+b hypothesis = X * W + b # cost/loss function cost = tf.reduce_mean(tf.square(hypothesis - y_train)) # Minimize train = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost) GradientDescent train = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost) Logistic Regression GradientDescent Cost? Model? (Hypothesis?) train = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost) Logistic Regression GradientDescent Cost? Model? (Hypothesis?) hypothesis = tf.sigmoid(tf.matmul(X, W) + b) # cost/loss function cost = -tf.reduce_mean(Y * tf.log(hypothesis) + (1 - Y) * tf.log(1 - hypothesis)) Softmax Classifier Cost? Model? (Hypothesis?) train = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost) GradientDescent logits = tf.matmul(X,W)+b hypothesis = tf.nn.softmax(logits) # Cross entropy cost/loss cost = tf.reduce_mean (-tf.reduce_sum(Y * tf.log(hypothesis), axis=1)) Softmax Classifier Cost? Model? (Hypothesis?) train = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost) GradientDescent logits = tf.matmul(X,W)+b hypothesis = tf.nn.softmax(logits) # Cross entropy cost/loss cost = tf.reduce_mean (-tf.reduce_sum(Y * tf.log(hypothesis), axis=1)) cost = tf.reduce_mean( tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=Y)) (Deep) Neural Nets NN for MNIST # input placeholders X = tf.placeholder(tf.float32, [None, 784]) Y = tf.placeholder(tf.float