Examples of Keras (start)

Example of Keras (start):

1 Multi-class Softmax based on multilayer perceptron:

From keras.models import sequential from
keras.layers import dense, dropout, activationfrom keras.optimizers import S GD
model = sequential ()
# Dense (a) is a fully-connected layer with a hidden units.
# in the first layer, you must specify the expected input data shape:
# here, 20-dimensional vectors.
Model.add (Dense (input_dim=20, init= ' uniform ')) Model.add (
Activation (' Tanh ')) Model.add (
Dropout (0.5))
Model.add (Dense (init= ' uniform ')) Model.add (Activation (
' Tanh ')) Model.add (
Dropout (0.5)
) Model.add (Dense (ten, init= ' uniform '))
Model.add (Activation (' Softmax '))

sgd = SGD (lr=0.1, decay=1e-6, momentum=0.9, Nesterov=true)
model.compile (loss= ' categorical_crossentropy ',
                           optimizer=sgd,
                           metrics=[' accuracy ')
model.fit (X_train, Y_train,
               nb_epoch=20,
               batch_size=16)
score = Model.evaluate (X_test, Y_test, batch_size=16)

2 Another implementation of similar MLP:

Model = sequential ()
model.add (Dense (input_dim=20, activation= ' Relu '))
Model.add (Dropout (0.5)
) Model.add (Dense (activation= ' relu ')) Model.add (Dropout
(0.5)) Model.add (
dense, activation= ' Softmax ' ))
model.compile (loss= ' categorical_crossentropy ',
                           optimizer= ' Adadelta ',
                           metrics=[' accuracy '])

3 Multilayer perceptron for two classification

Model = sequential ()
model.add (Dense (input_dim=20, init= ' uniform ', activation= ' Relu '))
Model.add ( Dropout (0.5))
Model.add (dense (, activation= ' Relu ')) Model.add (Dropout
(0.5))
Model.add (dense (1, activation= ' sigmoid '))
model.compile (loss= ' binary_crossentropy ',
                            optimizer= ' Rmsprop ',
                            metrics=[ ' Accuracy '])

4 convolutional neural Networks similar to VGG:

From keras.models import sequential

From keras.layers import dense, dropout, Activation, Flatten

From keras.layers import convolution2d, maxpooling2dfrom keras.optimizers import SGD
Model = sequential () # INPUT:100X100 images with 3 channels (3, +) tensors.# This applies convolution filt ERs of size 3x3 Each.model.add (convolution2d (3, 3, border_mode= ' valid ', input_shape= (3,))) Model.add ( Activation (' Relu ')) Model.add (convolution2d (32, 3, 3))
Model.add (Activation (' Relu ')) Model.add (Maxpooling2d (pool_size= (2, 2)) Model.add (Dropout (0.25))
Model.add (convolution2d (3, 3, border_mode= ' valid ')) Model.add (Activation (' Relu ')) Model.add (convolution2d (64, 3, 3)) Model.add (Activation (' Relu ')) Model.add (Maxpooling2d (pool_size= (2, 2)) Model.add (Dropout (0.25))
Model.add (Flatten ()) # Note:keras does automatic shape inference.model.add (dense) model.add (Activation (' Relu ')) Model.add (Dropout (0.5))
Model.add (dense (ten)) Model.add (Activation (' Softmax '))
SGD = SGD (lr=0.1, decay=1e-6, momentum=0.9, Nesterov=true) model.compile (loss= ' categorical_crossentropy ', optimizer= sgd
Model.fit (X_train, Y_train, batch_size=32, nb_epoch=1)

5 Sequence classification using LSTM

From Keras.models  import sequential from
keras.layers  import Dense, dropout, Activation
from Keras.layers  Import embedding from
keras.layers   import LSTM

model = sequential ()
Model.add ( Embedding (Max_features, Input_length=maxlen))
Model.add (LSTM (output_dim=128, activation= ' sigmoid ', inner_ activation= ' hard_sigmoid '))
Model.add (Dropout (0.5))
Model.add (dense (1))
Model.add (Activation (' Sigmoid ')

model.compile (loss= ' binary_crossentropy ',
                          optimizer= ' Rmsprop ',
                          metrics=[' accuracy '))

Model.fit (X_train, Y_train, batch_size=16, nb_epoch=10)
score = Model.evaluate (X_test, Y_test, batch_size= 16)

Use recursive units with thresholds for image description:

Note that for this network to work well, a larger volume of convolutional neural networks is required and initialized with pre-trained weights, which are only examples of structs.

Max_caption_len =
vocab_size = 10000 # First, let's

define an image model that # would
encode pictures into 128-dimensional vectors.
# It should is initialized with pre-trained weights.
Image_model = Sequential ()
Image_model.add (convolution2d (3, 3, border_mode= ' valid ', input_shape= (3, 100, 100)) )
Image_model.add (Activation (' Relu '))
Image_model.add (convolution2d (3, 3))
Image_model.add ( Activation (' Relu '))
Image_model.add (Maxpooling2d (pool_size= (2, 2))

Image_model.add (convolution2d (3, 3, border_mode= ' valid ')) Image_model.add (Activation (' Relu ')) Image_model.add ( Convolution2d (3, 3)) Image_model.add (Activation (' Relu ')) Image_model.add (Maxpooling2d (pool_size= (2, 2)) Image_
Model.add (Flatten ()) Image_model.add (Dense ()) # Let's load the weights from a save file. Image_model.load_weights (' Weight_file.h5 ') # Next, let's define a RNN model that encodes sequences of words # into Sequenc
Es of 128-dimensional word vectors. Language_model = Sequential () Language_model.add (Embedding (Vocab_size,, Input_length=max_caption_len)) Language_ Model.add (GRU (output_dim=128, Return_sequences=true)) Language_model.add (timedistributed (Dense) # Let's repeat
The image vector to turn it into a sequence. Image_model.add (Repeatvector (Max_caption_len)) # The output of both models would be tensors of shape (samples, max_caption_
Len, 128).
# let ' s concatenate these 2 vector sequences. Model = sequential () model.add (Merge ([Image_model, Language_model], mode= ' concat ', concat_axis=-1)) # Let's encode this vector sequence to a single vector model.add (GRU, return_sequ Ences=false) # which'll be used to compute a probability # distribution over what the next word in the caption should b
E! Model.add (Dense (vocab_size)) Model.add (Activation (' Softmax ')) model.compile (loss= ' categorical_crossentropy ',
Optimizer= ' Rmsprop ') # "Images" is a numpy float array of shape (Nb_samples, nb_channels=3, width, height). # "Captions" is a numpy integer array of shape (Nb_samples, Max_caption_len) # containing word index sequences representin
G Partial captions.  # "Next_words" is a numpy float array of shape (Nb_samples, vocab_size) # containing a categorical encoding (0s and 1s) of
The next word in the corresponding # partial caption. Model.fit ([Images, partial_captions], next_words, batch_size=16, nb_epoch=100)