In [1]:

    

%pwd


    

    
Out[1]:





'/Users/admin/Desktop/wasit/cs634/2017/cn350_project'

In [2]:

    

import tensorflow as tf


    

In [3]:

    

P = tf.Variable(tf.ones([3]),tf.float32)
Q = tf.Variable(tf.random_uniform(shape=[3],minval=0,maxval=1,dtype=tf.float32))
#R=tf.add(P,Q)
R= 0.5*(P+Q)


    

In [4]:

    

with tf.Session() as sess:
    sess.run( tf.global_variables_initializer() )
    #sess.run( fetches= [P, Q], feed_dict=None )
    p,q, r = sess.run( [P, Q, R])
print(p,q,r)


    

    




[ 1.  1.  1.] [ 0.22090411  0.48200917  0.40541804] [ 0.61045206  0.74100459  0.70270902]

In [5]:

    

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 [6]:

    

x  =tf.placeholder(tf.float32,shape=[None,784]) #28*28
w0  =tf.Variable(tf.zeros([784,200]),tf.float32)
b0  =tf.Variable(tf.zeros([200]),tf.float32)
w1  =tf.Variable(tf.zeros([200,10]),tf.float32)
b1  =tf.Variable(tf.zeros([10]),tf.float32)
yt =tf.placeholder(tf.float32,shape=[None,10])


    

In [7]:

    

z0=tf.nn.softmax( tf.matmul(x,w0)+b0 )
yp=tf.nn.sigmoid( tf.matmul(z0,w1)+b1 )
cross_entropy=tf.reduce_mean(-tf.reduce_sum(yt*tf.log(yp),reduction_indices=[1]))
#cross_entropy=tf.reduce_mean(tf.reduce_sum(tf.square(yt - yp),reduction_indices=[1]))
train_step=tf.train.AdadeltaOptimizer(0.5).minimize(cross_entropy)


    

In [8]:

    

with tf.Session() as sess:
    sess.run( tf.global_variables_initializer() )
    for i in range(10000):
        #batch[0]: images
        #batch[1]: labels
        batch=mnist.train.next_batch(100)
        _,cost =sess.run(fetches=[train_step, cross_entropy], 
                 feed_dict={ x: batch[0], yt: batch[1] })
        if i%1000==0:
            print(cost) 
    correct_prediction=tf.equal(tf.argmax(yp,1),tf.arg_max(yt,1))
    accuracy=tf.reduce_mean(tf.cast(correct_prediction,tf.float32))
    acc=accuracy.eval(feed_dict={x:mnist.test.images, yt:mnist.test.labels})*100
    print ("Accuracy of the ANN is {}".format(acc) )


    

    




0.693147
0.524198
0.38278
0.281024
0.204577
0.150241
0.110953
0.0820108
0.0605303
0.0450831
Accuracy of the ANN is 11.349999904632568

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]: