

In [45]:

    
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)









    



Extracting MNIST_data/train-images-idx3-ubyte.gz
Extracting MNIST_data/train-labels-idx1-ubyte.gz
Extracting MNIST_data/t10k-images-idx3-ubyte.gz
Extracting MNIST_data/t10k-labels-idx1-ubyte.gz



In [46]:

    
%matplotlib inline
import matplotlib.pyplot as plt
sample = mnist.train.images[0]
print(sample.shape)
reshaped_sample = sample.reshape((28,28))
print(reshaped_sample.shape)









    



(784,)
(28, 28)

Training Data Sample



In [47]:

    
plt.imshow(reshaped_sample, cmap = 'gray')









    Out[47]:





<matplotlib.image.AxesImage at 0x1f29c57b518>

Testing Data Sample



In [48]:

    
test_sample = mnist.test.images[0]
print(test_sample.shape)
reshaped_test_sample = test_sample.reshape((28,28))
print(reshaped_sample.shape)









    



(784,)
(28, 28)



In [49]:

    
plt.imshow(reshaped_test_sample, cmap = 'gray')









    Out[49]:





<matplotlib.image.AxesImage at 0x1f29c5d5898>

Main Code

Declare parameters



In [50]:

    
import tensorflow as tf
x = tf.placeholder(tf.float32, [None, 784])
W = tf.Variable(tf.zeros([784, 10]))
b = tf.Variable(tf.zeros([10]))
y = tf.matmul(x,W) + b

reduce_sum : adds the elements in the second dimension of y
reduce_mean : computes the mean over all the examples in the batch
GradientDescentOptimizer



In [51]:

    
y_ = tf.placeholder(tf.float32, [None, 10])

softmax_cross_entropy_with_logits(logits, labels, dim=-1, name=None) : computes softmax cross entropy between logits and labels.
Many classification operations : https://www.tensorflow.org/api_docs/python/nn/classification



In [52]:

    
learning_rate = 0.5
cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(y, y_))
train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(cross_entropy)



In [53]:

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

Train

This is called stochastic gradient descent method.



In [54]:

    
for _ in range(1000):
    batch_xs, batch_ys = mnist.train.next_batch(100)
    sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys})



In [55]:

    
correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))



In [56]:

    
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
print(sess.run(accuracy, feed_dict={x: mnist.test.images, y_: mnist.test.labels}))