Scribd
Upload
40 views

Keras Tutorial Cheatsheet

Uploaded by

ouiam ouhdifa
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
40 views

Keras Tutorial Cheatsheet

Uploaded by

ouiam ouhdifa
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

Keras Cheatsheet:

Python Deep Learning Tutorial

This cheatsheet will take you step-by-step through training a convolutional neural network in Python using the
famous MNIST dataset for handwritten digits classification. Our classifier will boast over 99% accuracy.

Keras is our recommended library for deep learning in Python, especially for beginners. Its minimalist, modular
approach makes it a breeze to get deep neural networks up and running.

To see the most up-to-date full tutorial, as well as installation instructions, visit the online tutorial at
elitedatascience.com.

SETUP Preprocess class labels


Make sure you have the following installed on your computer: Y_train = np_utils.to_categorical(y_train, 10)
Y_test = np_utils.to_categorical(y_test, 10)
• Python 2.7+ or Python 3
• SciPy with NumPy
• Matplotlib (Optional, recommended for exploratory analysis)
• Theano* Define model architecture
*note: TensorFlow is also supported (as an alternative to Theano), but we model = Sequential()
stick with Theano to keep it simple. The main difference is that you’ll need
to reshape the data slightly differently before feeding it to your network.
model.add(Convolution2D(32, 3, 3, activation=’relu’,
input_shape=(1,28,28)))
Import libraries and modules model.add(Convolution2D(32, 3, 3, activation=’relu’))
import numpy as np
model.add(MaxPooling2D(pool_size=(2,2)))
np.random.seed(123) # for reproducibility
model.add(Dropout(0.25))

from keras.models import Sequential


model.add(Flatten())
from keras.layers import Dense, Dropout, Activation, Flatten
model.add(Dense(128, activation=’relu’))
from keras.layers import Convolution2D, MaxPooling2D
model.add(Dropout(0.5))
from keras.utils import np_utils
model.add(Dense(10, activation=’softmax’))
from keras.datasets import mnist

Compile model
Load pre-shuffled MNIST data model.compile(loss=’categorical_crossentropy’,
into train and test sets optimizer=’adam’,
(X_train, y_train), (X_test, y_test) = mnist.load_data() metrics=[‘accuracy’])

Preprocess input data Fit model on training data


X_train = X_train.reshape(X_train.shape[0], 1, 28, 28)
model.fit(X_train, Y_train,
X_test = X_test.reshape(X_test.shape[0], 1, 28, 28)
batch_size=32, nb_epoch=10, verbose=1)
X_train = X_train.astype(‘float32’)
X_test = X_test.astype(‘float32’)
X_train /= 255
Evaluate model on test data
score = model.evaluate(X_test, Y_test, verbose=0)
X_test /= 255

To see the most up-to-date full tutorial, explanations, and additional context, visit the online tutorial at elitedatascience.com.
We also have plenty of other tutorials and guides.

ELITEDATASCIENCE.COM

You might also like