Deep Neural Network in Keras

In this notebook, we build a Deep Neural Network while discussing the theory we've learnt thus far.

Set seed for reproducibility



In [ ]:

    
import numpy as np
np.random.seed(42)

Load dependencies



In [ ]:

    
import keras
from keras.datasets import mnist
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import Dropout # new!
from keras.optimizers import SGD

Load data



In [ ]:

    
(X_train, y_train), (X_test, y_test) = mnist.load_data()

Preprocess data



In [ ]:

    
X_train = X_train.reshape(60000, 784).astype('float32')
X_test = X_test.reshape(10000, 784).astype('float32')



In [ ]:

    
X_train /= 255
X_test /= 255



In [ ]:

    
n_classes = 10
y_train = keras.utils.to_categorical(y_train, n_classes)
y_test = keras.utils.to_categorical(y_test, n_classes)

Design neural network architecture



In [ ]:

    
model = Sequential()
model.add(Dense(64, activation='sigmoid', input_shape=(784,)))
model.add(Dense(10, activation='softmax'))



In [ ]:

    
model.summary()

Configure model



In [ ]:

    
model.compile(loss='mean_squared_error', optimizer=SGD(lr=0.01), metrics=['accuracy'])

Train!



In [ ]:

    
model.fit(X_train, y_train, batch_size=128, epochs=200, verbose=1, validation_data=(X_test, y_test))



In [ ]: