Beginning with Deep Learning Using TensorFlow: A Beginners Guide to TensorFlow and Keras for Practicing Deep Learning Principles and Applications

by Mohan Kumar Silaparasetty

Length: 286 pages
Edition: 1
Language: English
Publisher: BPB Publications
Publication Date: 2022-02-09
ISBN-10: 9355510470
ISBN-13: 9789355510471
Sales Rank: #4274730 (See Top 100 Books)

4 ratings

Print Book Look Inside

A Practicing Guide to TensorFlow and Deep Learning

Key Features

Equipped with a necessary introduction to Deep Learning and AI.
Includes demos and templates to give your projects a good start.
Find more on the most important facets of AI, image, and speech recognition.

Description

This book begins with the configuration of an Anaconda development environment, essential for practicing the deep learning process. The basics of machine learning, which are needed for Deep Learning, are explained in this book.

TensorFlow is the industry-standard library for Deep Learning, and thereby, it is covered extensively with both versions, 1.x and 2.x. As neural networks are the heart of Deep Learning, the book explains them in great detail and systematic fashion, beginning with a single neuron and progressing through deep multilayer neural networks. The emphasis of this book is on the practical application of key concepts rather than going in-depth with them.

After establishing a firm basis in TensorFlow and Neural Networks, the book explains the concepts of image recognition using Convolutional Neural Networks (CNN), followed by speech recognition, and natural language processing (NLP). Additionally, this book discusses Transformers, the most recent advancement in NLP.

What you will learn

Create machine learning models for classification and regression.
Utilize TensorFlow 1.x to implement neural networks.
Work with the Keras API and TensorFlow 2.
Learn to design and train image categorization models.
Construct translation and Q & A apps using transformer-based language models.

Who this book is for

This book is intended for those new to Deep Learning who want to learn about its principles and applications. Before you begin, you’ll need to be familiar with Python.

Cover Page
Title Page
Copyright Page
Dedication Page
About the Author
About the Reviewer
Acknowledgement
Preface
Errata
Table of Contents
1. Introduction to Artificial Intelligence
    Structure
    Objective
    Brief history of artificial intelligence
        Classification of AI
        How did we reach here?
    AI adoption by industries
    Conclusion
    Points to remember
2. Machine Learning
    Introduction
    Structure
    Objectives
    Defining machine learning
        Supervised learning
    Setting up the environment
        Using Google Colab
        Setting up local environment in Python
            Prerequisite
        Regression algorithms
            Code demo
        Multilinear regression
        Logistic regression
        Decision tree
        Support vector machine (SVM)
            Unsupervised learning
    Conclusion
    Questions
3. TensorFlow Programming
    Introduction
    Structure
    Objective
    TensorFlow development environment
    Introducing TensorFlow
        Elements of TensorFlow program
            Constant
            Variable
            Placeholder
            Session
    Constants, variables, and placeholders
        Linear algebra with TensorFlow
        Optimizer
            Applying optimizer to solve simple mathematical problems
    Conclusion
    Questions
4. Neural Networks
    Introduction
    Structure
    Objective
    About Neural Networks
        Inputs
        Weights
        Bias
        Net input function (F)
        Activation function (G)
        MNIST
    MNIST—single layer multi-neuron model
        Multilayer Neural Network
        Multilayer binary classifier
    ReLu activation function
    Multilayer multiclass neural network
    Conclusion
    Questions
5. TensorFlow 2
    Introduction
    Structure
    Objective
    Installing TensorFlow 2
        Using Anaconda Navigator
        From Anaconda command prompt
            Google Colab
        What is new in TensorFlow 2?
        Kera API
    Classification with Iris data set
    Conclusion
    Points to remember
6. Image Recognition
    Introduction
    Structure
    Objective
    Introducing Convolutional Neural Networks (CNN)
        Convolution layer
        MNIST with CNN
        Binary image classification with Keras
    Multiclass image classification
    Load from data frame—binary
    Load from data frame—multiclass
        Save and restore models
    Pre-trained models
        Transfer learning
    Inference with Webcam images
    Object detection
    Conclusion
    Points to remember
7. Speech Recognition
    Introduction
    Structure
    Objective
    What is speech recognition?—Historical perspective
        Application of speech recognition
    Natural Language Processing (NLP)
        Word Embedding
            Language model
    Recurrent Neural Networks (RNN)
    Text classification
        Transformers
            Pre-trained transformer models
        BERT
    Machine language translation
    Q&A—SQUAD
    Conclusion
    Further reading
Index