Practicals: Deep Learning

Upgrade tensorflow to the latest version.

!pip install tensorflow --upgrade

See the version installed on your machine

import tensorflow as tf
print(tf.__version__)


For example, you may get the following value

2.0.0

In order to create the above neural network model, you can test the following code.

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D
from tensorflow.keras.layers import MaxPool2D
from tensorflow.keras.layers import Flatten
from tensorflow.keras.layers import Dense

# Creating a sequential model
model = Sequential()
model.add(Dense(4, activation='relu', input_shape=(3,)))
model.add(Dense(units=2, activation='softmax'))

# compiling the model
model.compile(loss='mse',
     optimizer='sgd',
     metrics=['accuracy']


In the above model, we use Stochastic gradient descent optimizer and mean square error as the loss calculator.

In the code below, we use a SGD optimizer using a learning rate of 0.01.

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D
from tensorflow.keras.layers import MaxPool2D
from tensorflow.keras.layers import Flatten
from tensorflow.keras.layers import Dense
from tensorflow.keras.optimizers import SGD


# Creating a sequential model
model = Sequential()
model.add(Dense(4, activation='relu', input_shape=(3,)))
model.add(Dense(units=2, activation='softmax'))

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


MNIST and Sequential model

Please run the following code https://github.com/keras-team/keras/blob/master/examples/mnist_mlp.py

Observe the different layers: 2 hidden layers and 2 dropouts.

MNIST and Convolutional neural networks

Please check and run the following code https://github.com/keras-team/keras/blob/master/examples/mnist_cnn.py

Observe the different layers: 2D convolutional network, dropout, maxpooling and flatten.

LSTM models

Example 1 (Time Series Prediction with LSTM models): Please check and run the following code https://machinelearningmastery.com/time-series-prediction-lstm-recurrent-neural-networks-python-keras/

Example 2 (Text generation using LSTM models): Please check and run the following code https://github.com/keras-team/keras/blob/master/examples/lstm_text_generation.py

Observe how sequences are generated in these examples.