Classifying Handwritten Digits

Handwritten Digits

Introduction

Requirements

Library Imports

from keras.datasets import mnist
from keras import models
from keras import layers
from keras.utils import to_categorical

Preparing Data

(train_images, train_labels), (test_images, test_labels) = mnist.load_data()
train_images = train_images.reshape((60000, 28 * 28))
train_images = train_images.astype('float32') / 255
test_images = test_images.reshape((10000, 28 * 28))
test_images = test_images.astype('float32') / 255
train_labels = to_categorical(train_labels)
test_labels = to_categorical(test_labels)

Design Network Architecture

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

Training Network

network.compile(optimizer='rmsprop',
loss='categorical_crossentropy',
metrics=['accuracy'])
network.fit(train_images, train_labels, epochs=5, batch_size=128)

Testing Network

test_loss, test_acc = network.evaluate(test_images, test_labels)
print("test_acc:", test_acc)
print("Test Accuracy: {:.0%}".format(test_acc))

The Output of Network

References

Computer Engineer & Scientist