In [1]:

    

from keras.datasets import mnist


    

    




Using TensorFlow backend.

In [2]:

    

(train_images, train_labels), (test_images, test_labels) = mnist.load_data()


    

In [3]:

    

train_images.shape


    

    
Out[3]:





(60000, 28, 28)

In [4]:

    

len(train_images)


    

    
Out[4]:





60000

In [7]:

    

train_labels


    

    
Out[7]:





array([5, 0, 4, ..., 5, 6, 8], dtype=uint8)

In [9]:

    

test_images.shape


    

    
Out[9]:





(10000, 28, 28)

In [10]:

    

len(test_images)


    

    
Out[10]:





10000

In [11]:

    

test_labels


    

    
Out[11]:





array([7, 2, 1, ..., 4, 5, 6], dtype=uint8)

In [12]:

    

from keras import models
from keras import layers


    

In [13]:

    

network = models.Sequential()
network.add(layers.Dense(512, activation = "relu", input_shape=(28 * 28,)))
network.add(layers.Dense(10, activation = "softmax"))


    

In [14]:

    

network.summary()


    

    




_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense_1 (Dense)              (None, 512)               401920    
_________________________________________________________________
dense_2 (Dense)              (None, 10)                5130      
=================================================================
Total params: 407,050
Trainable params: 407,050
Non-trainable params: 0
_________________________________________________________________

In [15]:

    

network.compile(optimizer = "rmsprop", loss = "categorical_crossentropy", metrics = ["accuracy"])


    

In [16]:

    

train_images = train_images.reshape(60000, 28 * 28)
test_images = test_images.reshape(10000, 28 * 28)

train_images = train_images.astype("float32") / 255
test_images = test_images.astype("float32") / 255


    

In [17]:

    

from keras.utils import to_categorical

train_labels = to_categorical(train_labels)
test_labels = to_categorical(test_labels)


    

In [18]:

    

network.fit(train_images, train_labels, epochs = 5, batch_size = 128)


    

    




Epoch 1/5
60000/60000 [==============================] - 7s - loss: 0.2571 - acc: 0.9250       
Epoch 2/5
60000/60000 [==============================] - 7s - loss: 0.1022 - acc: 0.9695     
Epoch 3/5
60000/60000 [==============================] - 7s - loss: 0.0695 - acc: 0.9790     
Epoch 4/5
60000/60000 [==============================] - 7s - loss: 0.0500 - acc: 0.9847     
Epoch 5/5
60000/60000 [==============================] - 7s - loss: 0.0382 - acc: 0.9886     






    
Out[18]:





<keras.callbacks.History at 0x7f092591e7b8>

In [19]:

    

test_loss, test_acc = network.evaluate(test_images, test_labels)
print("test accuracy", test_acc)


    

    




 9824/10000 [============================>.] - ETA: 0stest accuracy 0.9775

In [ ]: