MNIST Data Set - Basic Approach

In [28]:

    

import tensorflow as tf


    

In [29]:

    

from tensorflow.examples.tutorials.mnist import input_data


    

In [30]:

    

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 [31]:

    

type(mnist)


    

    
Out[31]:





tensorflow.contrib.learn.python.learn.datasets.base.Datasets

In [32]:

    

mnist.train.images


    

    
Out[32]:





array([[0., 0., 0., ..., 0., 0., 0.],
       [0., 0., 0., ..., 0., 0., 0.],
       [0., 0., 0., ..., 0., 0., 0.],
       ...,
       [0., 0., 0., ..., 0., 0., 0.],
       [0., 0., 0., ..., 0., 0., 0.],
       [0., 0., 0., ..., 0., 0., 0.]], dtype=float32)

In [33]:

    

mnist.train.num_examples


    

    
Out[33]:





55000

In [34]:

    

mnist.test.num_examples


    

    
Out[34]:





10000

In [35]:

    

mnist.validation.num_examples


    

    
Out[35]:





5000

In [36]:

    

import matplotlib.pyplot as plt
%matplotlib inline


    

In [37]:

    

mnist.train.images[1].shape


    

    
Out[37]:





(784,)

In [38]:

    

plt.imshow(mnist.train.images[1].reshape(28,28))


    

    
Out[38]:





<matplotlib.image.AxesImage at 0x2c5b21274a8>






    

In [39]:

    

plt.imshow(mnist.train.images[1].reshape(28,28),cmap='gist_gray')


    

    
Out[39]:





<matplotlib.image.AxesImage at 0x2c5b217d4a8>






    

In [40]:

    

mnist.train.images[1].max()


    

    
Out[40]:





1.0

In [41]:

    

plt.imshow(mnist.train.images[1].reshape(784,1))


    

    
Out[41]:





<matplotlib.image.AxesImage at 0x2c5b21d4400>






    

In [42]:

    

plt.imshow(mnist.train.images[1].reshape(784,1),cmap='gist_gray',aspect=0.02)


    

    
Out[42]:





<matplotlib.image.AxesImage at 0x2c5b221f8d0>






    

In [43]:

    

x = tf.placeholder(tf.float32,shape=[None,784])


    

In [44]:

    

# 10 because 0-9 possible numbers
W = tf.Variable(tf.zeros([784,10]))


    

In [45]:

    

b = tf.Variable(tf.zeros([10]))


    

In [46]:

    

# Create the Graph
y = tf.matmul(x,W) + b


    

In [47]:

    

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


    

In [48]:

    

# Cross Entropy


    

In [49]:

    

cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(labels=y_true, logits=y))


    

In [50]:

    

optimizer = tf.train.GradientDescentOptimizer(learning_rate=0.5)


    

In [51]:

    

train = optimizer.minimize(cross_entropy)


    

In [52]:

    

init = tf.global_variables_initializer()


    

In [53]:

    

with tf.Session() as sess:
    sess.run(init)
    
    # Train the model for 1000 steps on the training set
    # Using built in batch feeder from mnist for convenience
    
    for step in range(1000):
        
        batch_x , batch_y = mnist.train.next_batch(100)
        
        sess.run(train,feed_dict={x:batch_x,y_true:batch_y})
        
    # Test the Train Model
    matches = tf.equal(tf.argmax(y,1),tf.argmax(y_true,1))
    
    acc = tf.reduce_mean(tf.cast(matches,tf.float32))
    
    print(sess.run(acc,feed_dict={x:mnist.test.images,y_true:mnist.test.labels}))


    

    




0.9142