# -*- coding:utf-8 -*-
import numpy as np
import sklearn.preprocessing as prep
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
def xavier_init(fan_in, fan_out, constant = 1):
low = -constant * np.sqrt(6.0 / (fan_in + fan_out))
high = constant * np.sqrt(6.0 / (fan_in + fan_out))
return tf.random_uniform((fan_in, fan_out), minval=low, maxval=high, dtype = tf.float32)
class AdditiveGaussianNoiseAutoencoder(object):
def __init__(self, n_input, n_hidden, transfer_function=tf.nn.softplus, optimizer=tf.train.AdamOptimizer(), scale=0.1):
self.n_input = n_input
self.n_hidden = n_hidden
self.transfer = transfer_function
self.scale = tf.placeholder(tf.float32)
self.training_scale = scale
network_weights = self._initialize_weights()
self.weights = network_weights
self.x = tf.placeholder(tf.float32, [None, self.n_input])
self.hidden = self.transfer(tf.add(tf.matmul(
self.x + scale * tf.random_normal((n_input,)),
self.weights["w1"]), self.weights["b1"]
))
self.reconstruction = tf.add(tf.matmul(self.hidden, self.weights['w2']), self.weights['b2'])
self.cost = 0.5 * tf.reduce_sum(tf.pow(tf.subtract(self.reconstruction, self.x), 2.0))
self.optimiezer = optimizer.minimize(self.cost)
init = tf.global_variables_initializer()
self.sess = tf.Session()
self.sess.run(init)
def _initialize_weights(self):
all_weights = dict()
all_weights["w1"] = tf.Variable(xavier_init(self.n_input, self.n_hidden))
all_weights["b1"] = tf.Variable(tf.zeros([self.n_hidden], dtype = tf.float32))
all_weights["w2"] = tf.Variable(tf.zeros([self.n_hidden, self.n_input], dtype = tf.float32))
all_weights["b2"] = tf.Variable(tf.zeros([self.n_input], dtype = tf.float32))
return all_weights
def paritial_fit(self, X):
cost, opt = self.sess.run((self.cost, self.optimiezer),
feed_dict={self.x:X,self.scale:self.training_scale})
return cost
def calc_total_cost(self, X):
return self.sess.run(self.cost, feed_dict={self.x:X, self.scale:self.training_scale})
def transform(self, X):
return self.sess.run(self.hidden, feed_dict={self.x:X, self.scale:self.training_scale})
def generate(self, hidden = None):
if hidden is None:
hidden = np.random.normal(size=self.weights["b1"])
return self.sess.run(self.reconstruction,feed_dict={self.hidden:hidden})
def reconstruct(self, X):
return self.sess.run(self.reconstruction, feed_dict={self.x:X, self.scale:self.training_scale})
def getWeights(self):
return self.sess.run(self.weights["w1"])
def getBiases(self):
return self.sess.run(self.weights["b1"])
mnist = input_data.read_data_sets("MNIST_data", one_hot=True)
def standard_scale(X_train, X_test):
preprocessor = prep.StandardScaler().fit(X_train)
X_train = preprocessor.transform(X_train)
X_test = preprocessor.transform(X_test)
return X_train, X_test
def get_random_block_from_data(data, batch_size):
start_index = np.random.randint(0, len(data) - batch_size)
return data[start_index:(start_index + batch_size)]
X_train, X_test = standard_scale(mnist.train.images, mnist.test.images)
n_samples = int(mnist.train.num_examples)
training_epochs = 20
batch_size = 128
display_step = 1
autoencoder = AdditiveGaussianNoiseAutoencoder(n_input=784, n_hidden=200, transfer_function=tf.nn.softplus,
optimizer=tf.train.AdamOptimizer(learning_rate=0.001),scale=0.01)
for epoch in range(training_epochs):
avg_cost = 0
total_batch = int(n_samples / batch_size)
for i in range(total_batch):
batch_xs = get_random_block_from_data(X_train, batch_size)
cost = autoencoder.paritial_fit(batch_xs)
avg_cost += cost / n_samples * batch_size
if epoch % display_step == 0:
print("Epoch:", '%04d' %(epoch + 1), "cost=", "{:.9f}".format(avg_cost))
print("Total cost:" + str(autoencoder.calc_total_cost(X_test)))
Epoch: 0001 cost= 18638.615887500
Epoch: 0002 cost= 12971.389887500
Epoch: 0003 cost= 10659.026995455
Epoch: 0004 cost= 9836.657871023
Epoch: 0005 cost= 9484.297226705
Epoch: 0006 cost= 9271.624574432
Epoch: 0007 cost= 8899.998528409
Epoch: 0008 cost= 9324.328033523
Epoch: 0009 cost= 8969.404742045
Epoch: 0010 cost= 8170.966078409
Epoch: 0011 cost= 8705.253703977
Epoch: 0012 cost= 8673.962509659
Epoch: 0013 cost= 8482.184548864
Epoch: 0014 cost= 8119.183536932
Epoch: 0015 cost= 8492.594153409
Epoch: 0016 cost= 7844.617388636
Epoch: 0017 cost= 7948.973987500
Epoch: 0018 cost= 8431.035121591
Epoch: 0019 cost= 7783.220775568
Epoch: 0020 cost= 7945.380931250
Total cost:685076.2