Book Image

Deep Learning with Keras

By : Antonio Gulli, Sujit Pal
Book Image

Deep Learning with Keras

By: Antonio Gulli, Sujit Pal

Overview of this book

This book starts by introducing you to supervised learning algorithms such as simple linear regression, the classical multilayer perceptron and more sophisticated deep convolutional networks. You will also explore image processing with recognition of handwritten digit images, classification of images into different categories, and advanced objects recognition with related image annotations. An example of identification of salient points for face detection is also provided. Next you will be introduced to Recurrent Networks, which are optimized for processing sequence data such as text, audio or time series. Following that, you will learn about unsupervised learning algorithms such as Autoencoders and the very popular Generative Adversarial Networks (GANs). You will also explore non-traditional uses of neural networks as Style Transfer. Finally, you will look at reinforcement learning and its application to AI game playing, another popular direction of research and application of neural networks.

Related Content you might be interested in

Current Title:

Small book image
Deep Learning with Keras
Antonio Gulli | Sujit Pal
Apr, 2017 | 318 pages
Small book image
Deep Learning with TensorFlow 2 and Keras
Mr. Antonio Gulli | Dr. Amita Kapoor | Sujit Pal
Dec, 2019 | 646 pages
Small book image
Deep Learning with TensorFlow and Keras – 3rd edition
Dr. Amita Kapoor | Mr. Antonio Gulli | Sujit Pal
Oct, 2022 | 698 pages
Small book image
TensorFlow 1.x Deep Learning Cookbook
Sujit Pal | Dr. Amita Kapoor | Narotam Singh | Mr. Antonio Gulli
Dec, 2017 | 536 pages
Table of Contents (16 chapters)
Title Page
Title Page
Credits
Credits
About the Authors
About the Authors
About the Reviewer
About the Reviewer
www.PacktPub.com
www.PacktPub.com
Customer Feedback
Customer Feedback
Preface
Preface
Free Chapter
1
Neural Networks Foundations
Neural Networks Foundations
Perceptron
Multilayer perceptron — the first example of a network
A real example — recognizing handwritten digits
A practical overview of backpropagation
Towards a deep learning approach
Summary
2
Keras Installation and API
Keras Installation and API
Installing Keras
Configuring Keras
Installing Keras on Docker
Installing Keras on Google Cloud ML
Installing Keras on Amazon AWS
Installing Keras on Microsoft Azure
Keras API
Callbacks for customizing the training process
Summary
3
Deep Learning with ConvNets
Deep Learning with ConvNets
Deep convolutional neural network — DCNN
An example of DCNN — LeNet
Recognizing CIFAR-10 images with deep learning
Very deep convolutional networks for large-scale image recognition
Summary
4
Generative Adversarial Networks and WaveNet
Generative Adversarial Networks and WaveNet
What is a GAN?
Deep convolutional generative adversarial networks
Keras adversarial GANs for forging MNIST
Keras adversarial GANs for forging CIFAR
WaveNet — a generative model for learning how to produce audio
Summary
5
Word Embeddings
Word Embeddings
Distributed representations
word2vec
Exploring GloVe
Using pre-trained embeddings
Summary
6
Recurrent Neural Network — RNN
Recurrent Neural Network — RNN
SimpleRNN cells
RNN topologies
Vanishing and exploding gradients
Long short term memory — LSTM
Gated recurrent unit — GRU
Bidirectional RNNs
Stateful RNNs
Other RNN variants
Summary
7
Additional Deep Learning Models
Additional Deep Learning Models
Keras functional API
Regression networks
Unsupervised learning — autoencoders
Composing deep networks
Customizing Keras
Generative models
Summary
8
AI Game Playing
AI Game Playing
Reinforcement learning
Example - Keras deep Q-network for catch
The road ahead
Summary
9
Conclusion
Conclusion
Keras 2.0 — what is new

Deep convolutional neural network — DCNN

A deep convolutional neural network (DCNN) consists of many neural network layers. Two different types of layers, convolutional and pooling, are typically alternated. The depth of each filter increases from left to right in the network. The last stage is typically made of one or more fully connected layers:

There are three key intuitions beyond ConvNets:

  • Local receptive fields
  • Shared weights
  • Pooling

Let's review them.

Local receptive fields

If we want to preserve spatial information, then it is convenient to represent each image with a matrix of pixels. Then, a simple way to encode the local structure is to connect a submatrix of adjacent input neurons into one single hidden neuron belonging to the next layer. That single hidden neuron represents one local receptive field. Note that this operation is named convolution and it gives the name to this type of network.

Of course, we can encode more information by having overlapping submatrices. For instance, let...

Previous Section
End of Section 2
Next Section