100% found this document useful (1 vote)
52 views53 pages

Pro Deep Learning With Tensorflow 2.0: A Mathematical Approach To Advanced Artificial Intelligence in Python 2Nd Edition Santanu Pattanayak

The document promotes the book 'Pro Deep Learning with TensorFlow 2.0' by Santanu Pattanayak and provides links to download it along with other recommended titles related to artificial intelligence and deep learning. It includes details about the book's content, structure, and the author's background. Additionally, it mentions copyright information and the availability of supplementary materials on GitHub.

Uploaded by

nqobyacit
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
100% found this document useful (1 vote)
52 views53 pages

Pro Deep Learning With Tensorflow 2.0: A Mathematical Approach To Advanced Artificial Intelligence in Python 2Nd Edition Santanu Pattanayak

The document promotes the book 'Pro Deep Learning with TensorFlow 2.0' by Santanu Pattanayak and provides links to download it along with other recommended titles related to artificial intelligence and deep learning. It includes details about the book's content, structure, and the author's background. Additionally, it mentions copyright information and the availability of supplementary materials on GitHub.

Uploaded by

nqobyacit
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 53

Visit ebookmass.

com to download the full version and


explore more ebook or textbook

Pro Deep Learning with TensorFlow 2.0: A


Mathematical Approach to Advanced Artificial
Intelligence in Python 2nd Edition Santanu
Pattanayak
_____ Click the link below to download _____
https://ebookmass.com/product/pro-deep-learning-with-
tensorflow-2-0-a-mathematical-approach-to-advanced-
artificial-intelligence-in-python-2nd-edition-santanu-
pattanayak/

Explore and download more ebook or textbook at ebookmass.com


Here are some recommended products that we believe you will be
interested in. You can click the link to download.

Artificial Intelligence and Deep Learning in Pathology 1st


Edition Stanley Cohen Md (Editor)

https://ebookmass.com/product/artificial-intelligence-and-deep-
learning-in-pathology-1st-edition-stanley-cohen-md-editor/

Artificial Intelligence Programming with Python from Zero


to Hero 1st Edition Perry Xiao

https://ebookmass.com/product/artificial-intelligence-programming-
with-python-from-zero-to-hero-1st-edition-perry-xiao/

Risk Modeling: Practical Applications of Artificial


Intelligence, Machine Learning, and Deep Learning Terisa
Roberts
https://ebookmass.com/product/risk-modeling-practical-applications-of-
artificial-intelligence-machine-learning-and-deep-learning-terisa-
roberts/

Python for Artificial Intelligence. A Comprehensive Guide


Elsherif H.

https://ebookmass.com/product/python-for-artificial-intelligence-a-
comprehensive-guide-elsherif-h/
Artificial Intelligence: A Modern Approach 4th Edition
Stuart Russell

https://ebookmass.com/product/artificial-intelligence-a-modern-
approach-4th-edition-stuart-russell/

Practical MATLAB Deep Learning: A Projects-Based Approach,


2nd Edition Michael Paluszek

https://ebookmass.com/product/practical-matlab-deep-learning-a-
projects-based-approach-2nd-edition-michael-paluszek/

Artificial Intelligence for Dummies 2nd Edition John Paul


Mueller & Luca Massaron

https://ebookmass.com/product/artificial-intelligence-for-dummies-2nd-
edition-john-paul-mueller-luca-massaron/

Learning Scientific Programming with Python 2nd Edition


Christian Hill

https://ebookmass.com/product/learning-scientific-programming-with-
python-2nd-edition-christian-hill-2/

Learning Scientific Programming With Python 2nd Edition


Christian Hill

https://ebookmass.com/product/learning-scientific-programming-with-
python-2nd-edition-christian-hill/
Pro Deep
Learning with
TensorFlow 2.0
A Mathematical Approach to Advanced
Artificial Intelligence in Python

Second Edition

Santanu Pattanayak
Pro Deep Learning with
TensorFlow 2.0
A Mathematical Approach to
Advanced Artificial Intelligence
in Python
Second Edition

Santanu Pattanayak
Pro Deep Learning with TensorFlow 2.0: A Mathematical Approach to Advanced
Artificial Intelligence in Python
Santanu Pattanayak
Prestige Ozone
Bangalore, Karnataka, India

ISBN-13 (pbk): 978-1-4842-8930-3 ISBN-13 (electronic): 978-1-4842-8931-0


https://doi.org/10.1007/978-1-4842-8931-0
Copyright © 2023 by Santanu Pattanayak
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
Trademarked names, logos, and images may appear in this book. Rather than use a trademark symbol with
every occurrence of a trademarked name, logo, or image we use the names, logos, and images only in an
editorial fashion and to the benefit of the trademark owner, with no intention of infringement of the
trademark.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not
identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to
proprietary rights.
While the advice and information in this book are believed to be true and accurate at the date of publication,
neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or
omissions that may be made. The publisher makes no warranty, express or implied, with respect to the
material contained herein.
Managing Director, Apress Media LLC: Welmoed Spahr
Acquisitions Editor: Celestin Suresh John
Development Editor: Laura Berendson
Coordinating Editor: Aditee Mirashi
Cover designed by eStudioCalamar
Cover image by Pixabay (www.pixabay.com)
Distributed to the book trade worldwide by Apress Media, LLC, 1 New York Plaza, New York, NY 10004,
U.S.A. Phone 1-800-SPRINGER, fax (201) 348-4505, e-mail [email protected], or visit www.
springeronline.com. Apress Media, LLC is a California LLC and the sole member (owner) is Springer Science
+ Business Media Finance Inc (SSBM Finance Inc). SSBM Finance Inc is a Delaware corporation.
For information on translations, please e-mail [email protected]; for reprint,
paperback, or audio rights, please e-mail [email protected].
Apress titles may be purchased in bulk for academic, corporate, or promotional use. eBook versions and
licenses are also available for most titles. For more information, reference our Print and eBook Bulk Sales
web page at http://www.apress.com/bulk-sales.
Any source code or other supplementary material referenced by the author in this book is available to
readers on GitHub (https://github.com/Apress). For more detailed information, please visit http://www.
apress.com/source-code.
Printed on acid-free paper
To my wife, Sonia.
Table of Contents
About the Author��������������������������������������������������������������������������������������������������� xiii

About the Technical Reviewer���������������������������������������������������������������������������������xv

Introduction�����������������������������������������������������������������������������������������������������������xvii

Chapter 1: Mathematical Foundations��������������������������������������������������������������������� 1


Linear Algebra������������������������������������������������������������������������������������������������������������������������������� 2
Vector�������������������������������������������������������������������������������������������������������������������������������������� 3
Scalar�������������������������������������������������������������������������������������������������������������������������������������� 4
Matrix�������������������������������������������������������������������������������������������������������������������������������������� 4
Tensor�������������������������������������������������������������������������������������������������������������������������������������� 5
Matrix Operations and Manipulations�������������������������������������������������������������������������������������� 6
Linear Independence of Vectors�������������������������������������������������������������������������������������������� 10
Rank of a Matrix�������������������������������������������������������������������������������������������������������������������� 12
Identity Matrix or Operator���������������������������������������������������������������������������������������������������� 13
Determinant of a Matrix�������������������������������������������������������������������������������������������������������� 13
Inverse of a Matrix����������������������������������������������������������������������������������������������������������������� 16
Norm of a Vector�������������������������������������������������������������������������������������������������������������������� 17
Pseudo-Inverse of a Matrix��������������������������������������������������������������������������������������������������� 19
Unit Vector in the Direction of a Specific Vector�������������������������������������������������������������������� 19
Projection of a Vector in the Direction of Another Vector������������������������������������������������������� 20
Eigen Vectors������������������������������������������������������������������������������������������������������������������������� 21
Calculus�������������������������������������������������������������������������������������������������������������������������������������� 28
Differentiation������������������������������������������������������������������������������������������������������������������������ 28
Gradient of a Function����������������������������������������������������������������������������������������������������������� 29
Successive Partial Derivatives���������������������������������������������������������������������������������������������� 30

v
Table of Contents

Hessian Matrix of a Function������������������������������������������������������������������������������������������������� 30


Maxima and Minima of Functions����������������������������������������������������������������������������������������� 31
Local Minima and Global Minima������������������������������������������������������������������������������������������ 34
Positive Semi-definite and Positive Definite�������������������������������������������������������������������������� 35
Convex Set����������������������������������������������������������������������������������������������������������������������������� 35
Convex Function�������������������������������������������������������������������������������������������������������������������� 35
Non-convex Function������������������������������������������������������������������������������������������������������������� 36
Multivariate Convex and Non-convex Functions Examples��������������������������������������������������� 36
Taylor Series�������������������������������������������������������������������������������������������������������������������������� 39
Probability����������������������������������������������������������������������������������������������������������������������������������� 40
Unions, Intersection, and Conditional Probability������������������������������������������������������������������ 41
Chain Rule of Probability for Intersection of Event���������������������������������������������������������������� 43
Mutually Exclusive Events����������������������������������������������������������������������������������������������������� 43
Independence of Events�������������������������������������������������������������������������������������������������������� 43
Conditional Independence of Events������������������������������������������������������������������������������������� 44
Bayes Rule����������������������������������������������������������������������������������������������������������������������������� 44
Probability Mass Function����������������������������������������������������������������������������������������������������� 45
Probability Density Function�������������������������������������������������������������������������������������������������� 45
Expectation of a Random Variable����������������������������������������������������������������������������������������� 46
Variance of a Random Variable���������������������������������������������������������������������������������������������� 46
Skewness and Kurtosis��������������������������������������������������������������������������������������������������������� 47
Covariance����������������������������������������������������������������������������������������������������������������������������� 50
Correlation Coefficient����������������������������������������������������������������������������������������������������������� 50
Some Common Probability Distribution��������������������������������������������������������������������������������� 51
Likelihood Function��������������������������������������������������������������������������������������������������������������� 60
Maximum Likelihood Estimate���������������������������������������������������������������������������������������������� 61
Hypothesis Testing and p Value��������������������������������������������������������������������������������������������� 63
Formulation of Machine-Learning Algorithm and Optimization Techniques������������������������������� 66
Supervised Learning�������������������������������������������������������������������������������������������������������������� 67
Unsupervised Learning���������������������������������������������������������������������������������������������������������� 79
Reinforcement Learning�������������������������������������������������������������������������������������������������������� 80

vi
Table of Contents

Optimization Techniques for Machine-Learning Gradient Descent���������������������������������������� 81


Constrained Optimization Problem���������������������������������������������������������������������������������������� 93
A Few Important Topics in Machine Learning����������������������������������������������������������������������������� 96
Dimensionality-Reduction Methods�������������������������������������������������������������������������������������� 96
Regularization��������������������������������������������������������������������������������������������������������������������� 102
Regularization Viewed as a Constraint Optimization Problem��������������������������������������������� 105
Bias and Variance Trade-Off������������������������������������������������������������������������������������������������ 106
Summary���������������������������������������������������������������������������������������������������������������������������������� 108

Chapter 2: Introduction to Deep-­Learning Concepts and TensorFlow������������������ 109


Deep Learning and Its Evolution����������������������������������������������������������������������������������������������� 109
Perceptrons and Perceptron Learning Algorithm���������������������������������������������������������������������� 113
Geometrical Interpretation of Perceptron Learning������������������������������������������������������������� 117
Limitations of Perceptron Learning������������������������������������������������������������������������������������� 118
Need for Nonlinearity����������������������������������������������������������������������������������������������������������� 121
Hidden-Layer Perceptrons’ Activation Function for Nonlinearity����������������������������������������� 123
Different Activation Functions for a Neuron/Perceptron������������������������������������������������������ 124
Learning Rule for Multi-layer Perceptron Network�������������������������������������������������������������� 134
Backpropagation for Gradient Computation������������������������������������������������������������������������ 136
Generalizing the Backpropagation Method for Gradient Computation�������������������������������� 138
TensorFlow�������������������������������������������������������������������������������������������������������������������������������� 146
Common Deep-Learning Packages������������������������������������������������������������������������������������� 147
TensorFlow Installation�������������������������������������������������������������������������������������������������������� 149
TensorFlow Basics for Development����������������������������������������������������������������������������������� 149
Gradient-Descent Optimization Methods from a Deep-­Learning Perspective��������������������� 157
Learning Rate in Mini-Batch Approach to Stochastic Gradient Descent����������������������������� 164
Optimizers in TensorFlow���������������������������������������������������������������������������������������������������� 164
XOR Implementation Using TensorFlow������������������������������������������������������������������������������� 176
Linear Regression in TensorFlow����������������������������������������������������������������������������������������� 184
Multiclass Classification with SoftMax Function Using Full-Batch Gradient Descent��������� 187
Multiclass Classification with SoftMax Function Using Stochastic Gradient Descent��������� 191

vii
Table of Contents

GPU������������������������������������������������������������������������������������������������������������������������������������������� 196
TPU�������������������������������������������������������������������������������������������������������������������������������������������� 196
Summary���������������������������������������������������������������������������������������������������������������������������������� 197

Chapter 3: Convolutional Neural Networks����������������������������������������������������������� 199


Convolution Operation��������������������������������������������������������������������������������������������������������������� 199
Linear Time Invariant (LTI)/Linear Shift Invariant (LSI) Systems������������������������������������������ 200
Convolution for Signals in One Dimension�������������������������������������������������������������������������� 202
Analog and Digital Signals�������������������������������������������������������������������������������������������������������� 205
2D and 3D Signals��������������������������������������������������������������������������������������������������������������� 207
2D Convolution�������������������������������������������������������������������������������������������������������������������������� 209
Two-Dimensional Unit Step Function���������������������������������������������������������������������������������� 209
2D Convolution of a Signal with an LSI System Unit Step Response����������������������������������� 211
2D Convolution of an Image to Different LSI System Responses���������������������������������������� 214
Common Image-Processing Filters������������������������������������������������������������������������������������������� 219
Mean Filter�������������������������������������������������������������������������������������������������������������������������� 220
Median Filter����������������������������������������������������������������������������������������������������������������������� 222
Gaussian Filter��������������������������������������������������������������������������������������������������������������������� 225
Gradient-Based Filters��������������������������������������������������������������������������������������������������������� 226
Sobel Edge-Detection Filter������������������������������������������������������������������������������������������������� 228
Identity Transform���������������������������������������������������������������������������������������������������������������� 231
Convolution Neural Networks��������������������������������������������������������������������������������������������������� 232
Components of Convolution Neural Networks�������������������������������������������������������������������������� 233
Input Layer��������������������������������������������������������������������������������������������������������������������������� 235
Convolution Layer���������������������������������������������������������������������������������������������������������������� 235
Pooling Layer����������������������������������������������������������������������������������������������������������������������� 238
Backpropagation Through the Convolutional Layer������������������������������������������������������������������ 239
Backpropagation Through the Pooling Layers�������������������������������������������������������������������������� 243
Weight Sharing Through Convolution and Its Advantages�������������������������������������������������������� 244
Translation Equivariance����������������������������������������������������������������������������������������������������������� 245
Translation Invariance Due to Pooling��������������������������������������������������������������������������������������� 247

viii
Table of Contents

Dropout Layers and Regularization������������������������������������������������������������������������������������������� 248


Convolutional Neural Network for Digit Recognition on the MNIST Dataset����������������������������� 250
Convolutional Neural Network for Solving Real-World Problems���������������������������������������������� 255
Batch Normalization������������������������������������������������������������������������������������������������������������������ 265
Different Architectures in Convolutional Neural Networks�������������������������������������������������������� 268
LeNet����������������������������������������������������������������������������������������������������������������������������������� 268
AlexNet�������������������������������������������������������������������������������������������������������������������������������� 270
VGG16���������������������������������������������������������������������������������������������������������������������������������� 272
ResNet��������������������������������������������������������������������������������������������������������������������������������� 273
Transfer Learning���������������������������������������������������������������������������������������������������������������������� 274
Guidelines for Using Transfer Learning�������������������������������������������������������������������������������� 276
Transfer Learning with Google’s InceptionV3���������������������������������������������������������������������� 277
Transfer Learning with Pretrained VGG16���������������������������������������������������������������������������� 278
Dilated Convolution������������������������������������������������������������������������������������������������������������������� 284
Depthwise Separable Convolution�������������������������������������������������������������������������������������������� 285
Summary���������������������������������������������������������������������������������������������������������������������������������� 290

Chapter 4: Natural Language Processing������������������������������������������������������������� 293


Vector Space Model (VSM)�������������������������������������������������������������������������������������������������������� 293
Vector Representation of Words������������������������������������������������������������������������������������������������ 298
Word2Vec���������������������������������������������������������������������������������������������������������������������������������� 299
Continuous Bag of Words (CBOW)��������������������������������������������������������������������������������������� 300
Continuous Bag of Words Implementation in TensorFlow��������������������������������������������������� 304
Skip-Gram Model for Word Embedding������������������������������������������������������������������������������� 310
Skip-Gram Implementation in TensorFlow�������������������������������������������������������������������������� 313
Global Co-occurrence Statistics-Based Word Vectors��������������������������������������������������������� 319
GloVe����������������������������������������������������������������������������������������������������������������������������������� 326
Word Analogy with Word Vectors����������������������������������������������������������������������������������������� 330
Introduction to Recurrent Neural Networks������������������������������������������������������������������������������ 334
Language Modeling������������������������������������������������������������������������������������������������������������� 338
Predicting the Next Word in a Sentence Through RNN Versus Traditional Methods������������ 339
Backpropagation Through Time (BPTT)������������������������������������������������������������������������������� 340

ix
Table of Contents

Vanishing- and Exploding-Gradient Problem in RNN����������������������������������������������������������� 344


Solution to Vanishing- and Exploding-Gradient Problem in RNNs��������������������������������������� 347
Long Short-Term Memory (LSTM)��������������������������������������������������������������������������������������� 348
LSTM in Reducing Exploding- and Vanishing-Gradient Problems�������������������������������������� 351
MNIST Digit Identification in TensorFlow Using Recurrent Neural Networks���������������������� 353
Gated Recurrent Unit (GRU)������������������������������������������������������������������������������������������������� 365
Bidirectional RNN���������������������������������������������������������������������������������������������������������������� 368
Architecture of the Neural Machine Translation Model Using Seq2Seq������������������������������ 370
Limitation of the Seq2Seq Model for Machine Translation�������������������������������������������������� 382
Attention����������������������������������������������������������������������������������������������������������������������������������� 383
Scaled Dot Product Attention���������������������������������������������������������������������������������������������������� 384
Multihead Attention������������������������������������������������������������������������������������������������������������������� 387
Transformers����������������������������������������������������������������������������������������������������������������������������� 387
Encoder������������������������������������������������������������������������������������������������������������������������������� 388
Decoder������������������������������������������������������������������������������������������������������������������������������� 389
Positional Encoding������������������������������������������������������������������������������������������������������������� 390
Final Output������������������������������������������������������������������������������������������������������������������������� 390
Summary���������������������������������������������������������������������������������������������������������������������������������� 404

Chapter 5: Unsupervised Learning with Restricted Boltzmann Machines and


Autoencoders������������������������������������������������������������������������������������������������������� 407
Boltzmann Distribution������������������������������������������������������������������������������������������������������������� 408
Bayesian Inference: Likelihood, Priors, and Posterior Probability Distribution�������������������������� 410
Markov Chain Monte Carlo Methods for Sampling�������������������������������������������������������������������� 416
Metropolis Algorithm����������������������������������������������������������������������������������������������������������� 420
Restricted Boltzmann Machines����������������������������������������������������������������������������������������������� 427
Training a Restricted Boltzmann Machine��������������������������������������������������������������������������� 434
Gibbs Sampling������������������������������������������������������������������������������������������������������������������� 440
Block Gibbs Sampling���������������������������������������������������������������������������������������������������������� 442
Burn-in Period and Generating Samples in Gibbs Sampling����������������������������������������������� 442
Using Gibbs Sampling in Restricted Boltzmann Machines�������������������������������������������������� 443
Contrastive Divergence������������������������������������������������������������������������������������������������������� 445

x
Visit https://ebookmass.com today to explore
a vast collection of ebooks across various
genres, available in popular formats like
PDF, EPUB, and MOBI, fully compatible with
all devices. Enjoy a seamless reading
experience and effortlessly download high-
quality materials in just a few simple steps.
Plus, don’t miss out on exciting offers that
let you access a wealth of knowledge at the
best prices!
Table of Contents

A Restricted Boltzmann Implementation in TensorFlow������������������������������������������������������ 446


Collaborative Filtering Using Restricted Boltzmann Machines�������������������������������������������� 452
Deep-Belief Networks (DBNs)���������������������������������������������������������������������������������������������� 458
Autoencoders���������������������������������������������������������������������������������������������������������������������������� 467
Feature Learning Through Autoencoders for Supervised Learning������������������������������������� 471
Kullback-Leibler (KL) Divergence���������������������������������������������������������������������������������������� 473
Sparse Autoencoder Implementation in TensorFlow����������������������������������������������������������� 477
Denoising Autoencoder�������������������������������������������������������������������������������������������������������� 482
A Denoising Autoencoder Implementation in TensorFlow��������������������������������������������������� 482
Variational Inference����������������������������������������������������������������������������������������������������������� 491
Variational Autoencoder Objective from ELBO��������������������������������������������������������������������� 494
Implementation Details of the Variational Autoencoder������������������������������������������������������ 496
Implementation of Variational Autoencoder������������������������������������������������������������������������� 498
PCA and ZCA Whitening������������������������������������������������������������������������������������������������������������ 505
Summary���������������������������������������������������������������������������������������������������������������������������������� 509

Chapter 6: Advanced Neural Networks����������������������������������������������������������������� 511


Image Segmentation����������������������������������������������������������������������������������������������������������������� 512
Binary Thresholding Method Based on Histogram of Pixel Intensities�������������������������������� 512
Otsu’s Method��������������������������������������������������������������������������������������������������������������������� 513
Watershed Algorithm for Image Segmentation�������������������������������������������������������������������� 518
Image Segmentation Using K-means Clustering����������������������������������������������������������������� 522
Semantic Segmentation������������������������������������������������������������������������������������������������������ 526
Sliding-Window Approach��������������������������������������������������������������������������������������������������� 527
Fully Convolutional Network (FCN)�������������������������������������������������������������������������������������� 528
Fully Convolutional Network with Downsampling and Upsampling������������������������������������ 531
U-Net����������������������������������������������������������������������������������������������������������������������������������� 537
Semantic Segmentation in TensorFlow with Fully Connected Neural Networks����������������� 539
Image Classification and Localization Network������������������������������������������������������������������������ 547
Object Detection������������������������������������������������������������������������������������������������������������������������ 549
R-CNN���������������������������������������������������������������������������������������������������������������������������������� 551
Fast and Faster R-CNN�������������������������������������������������������������������������������������������������������� 553

xi
Table of Contents

Generative Adversarial Networks���������������������������������������������������������������������������������������������� 554


Maximin and Minimax Problem������������������������������������������������������������������������������������������� 556
Zero-Sum Game������������������������������������������������������������������������������������������������������������������ 558
Minimax and Saddle Points������������������������������������������������������������������������������������������������� 560
GAN Cost Function and Training������������������������������������������������������������������������������������������ 562
Vanishing Gradient for the Generator���������������������������������������������������������������������������������� 565
GAN Learning from an F-Divergence Perspective��������������������������������������������������������������� 566
TensorFlow Implementation of a GAN Network������������������������������������������������������������������� 568
CycleGAN Implementation in TensorFlow���������������������������������������������������������������������������� 577
Geometric Deep Learning and Graph Neural Networks������������������������������������������������������������ 589
Manifolds����������������������������������������������������������������������������������������������������������������������������� 590
Implementation of Graph Classification Using GCN������������������������������������������������������������� 620
Summary���������������������������������������������������������������������������������������������������������������������������������� 634

Index��������������������������������������������������������������������������������������������������������������������� 635

xii
About the Author
Santanu Pattanayak works as a Senior Staff Machine
Learning Specialist at Qualcomm Corp R&D and is the
author of Quantum Machine Learning with Python,
published by Apress. He has more than 16 years of
experience, having worked at GE, Capgemini, and IBM
before joining Qualcomm. He graduated with a degree in
electrical engineering from Jadavpur University, Kolkata, and
is an avid math enthusiast. Santanu has a master’s degree in
data science from the Indian Institute of Technology (IIT),
Hyderabad. Currently, he resides in Bangalore with his wife.

xiii
About the Technical Reviewer
Manohar Swamynathan is a data science practitioner and
an avid programmer, with over 14+ years of experience
in various data science–related areas that include data
warehousing, business intelligence (BI), analytical tool
development, ad hoc analysis, predictive modeling, data
science product development, consulting, formulating
strategies, and executing analytics programs. He’s had
a career covering the life cycle of data across different
domains, such as US mortgage banking, retail/ecommerce,
insurance, and industrial IoT. He has a bachelor’s degree
with a specialization in physics, mathematics, and
computers and a master’s degree in project management.
He’s currently living in Bengaluru, the Silicon Valley of India.

xv
Introduction
Pro Deep Learning with TensorFlow 2.0 is a practical and mathematical guide to deep
learning using TensorFlow. Deep learning is a branch of machine learning where
you model the world in terms of a hierarchy of concepts. This pattern of learning is
similar to the way a human brain learns, and it allows computers to model complex
concepts that often go unnoticed in other traditional methods of modeling. Hence, in
the modern computing paradigm, deep learning plays a vital role in modeling complex
real-world problems, especially by leveraging the massive amount of unstructured data
available today.
Because of the complexities involved in a deep-learning model, many times it is
treated as a black box by people using it. However, to derive the maximum benefit from
this branch of machine learning, one needs to uncover the hidden mystery by looking
at the science and mathematics associated with it. In this book, great care has been
taken to explain the concepts and techniques associated with deep learning from a
mathematical as well as a scientific viewpoint. Also, the first chapter is totally dedicated
toward building the mathematical base required to comprehend deep-learning concepts
with ease. TensorFlow has been chosen as the deep-learning package because of
its flexibility for research purposes and its ease of use. Another reason for choosing
TensorFlow is its capability to load models with ease in a live production environment
using its serving capabilities.
In summary, Pro Deep Learning with TensorFlow 2.0 provides practical, hands-on
expertise so you can learn deep learning from scratch and deploy meaningful deep-­
learning solutions. This book will allow you to get up to speed quickly using TensorFlow
and to optimize different deep-learning architectures. All the practical aspects of deep
learning that are relevant in any industry are emphasized in this book. You will be able
to use the prototypes demonstrated to build new deep-learning applications. The code
presented in the book is available in the form of iPython notebooks and scripts that allow
you to try out examples and extend them in interesting ways. You will be equipped with
the mathematical foundation and scientific knowledge to pursue research in this field
and give back to the community.

xvii
Introduction

Who This Book Is For


This book is for data scientists and machine-learning professionals looking at
deep-­learning solutions to solve complex business problems.
This book is for software developers working on deep-learning solutions through
TensorFlow.
This book is for graduate students and open source enthusiasts with a constant
desire to learn.

What You’ll Learn


The chapters covered in this book are as follows:
Chapter 1—Mathematical Foundations: In this chapter, all the relevant mathematical
concepts from linear algebra, probability, calculus, optimization, and machine-learning
formulation are discussed in detail to lay the mathematical foundation required for deep
learning. The various concepts are explained with a focus on their use in the fields of
machine learning and deep learning.
Chapter 2—Introduction to Deep-Learning Concepts and TensorFlow: This chapter
introduces the world of deep learning and discusses its evolution over the years. The key
building blocks of neural networks, along with several methods of learning, such as the
Perceptron learning rule and backpropagation methods, are discussed in detail. Also, this
chapter introduces the paradigm of TensorFlow coding so that readers are accustomed to
the basic syntax before moving on to more involved implementations in TensorFlow.
Chapter 3—Convolutional Neural Networks: This chapter deals with convolutional
neural networks used for image processing. Image processing is a computer vision
issue that has seen a huge boost in performance in the areas of object recognition and
detection, object classification, localization, and segmentation using convolutional
neural networks. The chapter starts by illustrating the operation of convolution in detail
and then moves on to the working principles of a convolutional neural network. Much
emphasis is given to the building blocks of a convolutional neural network to give the
reader the tools needed to experiment and extend their networks in interesting ways.
Further, backpropagation through convolutional and pooling layers is discussed in
detail so that the reader has a holistic view of the training process of convolutional
networks. Also covered in this chapter are the properties of equivariance and translation
invariance, which are central to the success of convolutional neural networks.

xviii
Introduction

Chapter 4—Natural Language Processing: This chapter deals with natural language
processing using deep learning. It starts with different vector space models for text
processing; word-to-vector embedding models, such as the continuous bag of words
method and Skip-grams; and then moves to much more advanced topics that involve
recurrent neural networks (RNNs), LSTM, bidirectional RNN, and GRU. Language
modeling as well as neural machine translation is covered in detail in this chapter to
help the reader utilize these networks in real-world problems involving the same. Also,
the mechanism of backpropagation in cases of RNNs and LSTM as well vanishing-­
gradient problems are discussed in much detail. This chapter also puts emphasis on the
attention mechanism and how it fits into the powerful transformer architecture that has
changed the world of natural language processing.
Chapter 5—Unsupervised Learning with Restricted Boltzmann Machines and
Autoencoders: In this chapter, you will learn about unsupervised methods in deep
learning that use restricted Boltzmann machines (RBMs) and autoencoders. Also, the
chapter will touch upon Bayesian inference and Markov chain Monte Carlo (MCMC)
methods, such as the Metropolis algorithm and Gibbs sampling, since the RBM training
process requires some knowledge of sampling. Further, this chapter introduces the
concepts of contrastive divergence in the context of a customized version of Gibbs
sampling for enabling practical training of RBMs. We will further discuss how RBMs
can be used for collaborative filtering in recommender systems as well as their use in
unsupervised pretraining of deep-belief networks (DBNs).
In the second part of the chapter, various kinds of autoencoders are covered, such as
sparse encoders, denoising autoencoders, variational autoencoders, and so forth. Also,
the reader will learn about how internal features learned from the autoencoders can
be utilized for dimensionality reduction as well as for supervised learning. Finally, the
chapter ends with a little brief on data preprocessing techniques, such as PCA whitening
and ZCA whitening.
Chapter 6—Advanced Neural Networks: In this chapter, the reader will learn about
some of the advanced neural networks, such as fully convolutional neural networks,
R-CNN, Fast R-CNN, Faster, U-Net, and so forth, that deal with semantic segmentation
of images, object detection, and localization. This chapter also introduces the readers
to traditional image segmentation methods so that they can combine the best of both
worlds as appropriate. In the second half of the chapter, the reader will learn about
the generative adversarial network (GAN), a new schema of generative model used for
producing synthetic data like the data produced by a given distribution. GAN has usages

xix
Introduction

and potential in several fields, such as in image generation, image inpainting, abstract
reasoning, semantic segmentation, video generation, style transfer from one domain
to another, and text-to-image generation applications, among others. In addition to
standard GAN, the book introduces the theoretical concepts as well as the practical
deployment intricacies associated with cycle consistency GAN that is used for domain
translational problems.
To summarize, the key learnings the reader can expect from this book are as follows:

• Understand full-stack deep learning using TensorFlow and gain a


solid mathematical foundation for deep learning.

• Deploy complex deep-learning solutions in production using


TensorFlow.

• Carry out research on deep learning and perform experiments using


TensorFlow.

Source Code
All source code used in this book can be downloaded from github.com/apress/pro-
deep-learning-tensorflow2.

xx
CHAPTER 1

Mathematical Foundations
Deep learning is a branch of machine learning that uses many layers of artificial
neurons stacked one on top of the other for identifying complex features within the
input data and solving complex real-world problems. It can be used for both supervised
and unsupervised machine-learning tasks. Deep learning is currently used in areas
such as computer vision, video analytics, pattern recognition, anomaly detection, text
processing, sentiment analysis, and recommender system, among other things. Also, it
has widespread use in robotics, self-driving car mechanisms, and artificial intelligence
systems in general.
Mathematics is at the heart of any machine-learning algorithm. A strong grasp
of the core concepts of mathematics goes a long way in enabling one to select the
right algorithms for a specific machine-learning problem, keeping in mind the end
objectives. Also, it enables one to tune machine-learning/deep-learning models better
and understand the possible reasons for an algorithm not performing as desired.
Deep learning being a branch of machine learning demands as much expertise
in mathematics, if not more, than that required for other machine-learning tasks.
Mathematics as a subject is vast, but there are a few specific topics that machine-­
learning or deep-learning professionals and/or enthusiasts should be aware of to extract
the most out of this wonderful domain of machine learning, deep learning, and artificial
intelligence. Illustrated in Figure 1-1 are the different branches of mathematics along
with their importance in the field of machine learning and deep learning. We will discuss
the relevant concepts in each of the following branches in this chapter:

• Linear algebra

• Calculus

• Probability

• Optimization and formulation of machine-learning algorithms

1
© Santanu Pattanayak 2023
S. Pattanayak, Pro Deep Learning with TensorFlow 2.0, https://doi.org/10.1007/978-1-4842-8931-0_1
Chapter 1 Mathematical Foundations

Figure 1-1. Importance of mathematics topics for machine learning and


data science

Note Readers who are already familiar with these topics can choose to skip this
chapter or have a casual glance through the content.

Linear Algebra
Linear algebra is a branch of mathematics that deals with vectors and their
transformation from one vector space to another vector space. Since in machine
learning and deep learning we deal with multidimensional data and their manipulation,
linear algebra plays a crucial role in almost every machine-learning and deep-learning
algorithm. Illustrated in Figure 1-2 is a three-dimensional vector space where v1, v2, and
v3 are vectors and P is a 2-D plane within the three-dimensional vector space.

2
Visit https://ebookmass.com today to explore
a vast collection of ebooks across various
genres, available in popular formats like
PDF, EPUB, and MOBI, fully compatible with
all devices. Enjoy a seamless reading
experience and effortlessly download high-
quality materials in just a few simple steps.
Plus, don’t miss out on exciting offers that
let you access a wealth of knowledge at the
best prices!
Other documents randomly have
different content
credited herself with unusual powers of divination. She despised
other people for taking the world and its creatures at face-value.
“The amount of reading he must have done for this book is
enormous,” went on Littleton. “Because, unlike most wildly
enthusiastic reformers, who fling adjectives about and scream at the
top of their voices, he has marshalled an amazing array of facts and
figures. That, and his own discrimination and judgment, make the
book so fine. And there are one or two passages, where he lets
himself go, that are absolutely stirring. As you know”—with a laugh
—“I’m in the trade, and I don’t often enthuse over a book, but I was
greatly struck with this.”
“I am glad,” said Claudia dully, “very glad.”
This book had been in his mind for years, perhaps ever since he
left Oxford, and he had never talked of it to her. She would never
forgive him! He had not thought her worthy of his confidence. He
was not her friend. Then a vision of him at Fay’s flat that awful
night, quietly directing everyone and watching over her, came across
her mental vision, but this only confused her the more. Did he, like
most men, look upon her as a graceful, pretty plaything—just a
woman? Was his idea of a woman just like her husband’s, only
different in kind? Apparently she was of no real use to anyone,
except—yes, except to the little music-hall artiste whom the family
had rejected.
Then she looked at the man in the chair beside hers, and as her
preoccupation had made him drop his guard, she read clearly the
very personal admiration in his eyes. For a moment they remained
looking at one another, love in the man’s eyes, a hopeless
bewilderment and weariness in Claudia’s.
“Your life does not satisfy you,” said the man abruptly. “I have
known that for some time.”
“Is anyone satisfied with his life?”
She was a little startled, but a beautiful, much-sought-after
woman is seldom nonplussed by such a situation. She had seen that
look in too many men’s eyes. It was only startling with Littleton
because she had not noticed that he was falling in love with her. Was
that because she had been thinking of Frank to the exclusion of
other men? For though love itself may not be blind, it makes a
woman insensible to the feelings of other men and her very
preoccupation often piques them into desiring her.
Littleton got up and leaned against the mantelpiece, looking down
upon her. His straight, spare figure, in his unmistakable American
clothing, bespoke energy and endurance. The shape of his head, on
the forehead of which the fair hair was thinning a little, told of great
mental activity and powers of organization. Some woman might be
proud of such a man. In some ways he was not unlike Colin Paton,
save that he had the American restlessness and nerviness, and that
he lacked the fine polish and self-possession which a man may
possibly acquire, but is usually associated with families that can
count back many centuries, and that have always tried to uphold the
best traditions of English manhood. Paton’s ancestors had mainly
been divided into two classes, fighters and scholars. Admiral Worral
Paton had fought many a fight with Francis Drake on the high seas,
and another Paton in the reign of Elizabeth had been accounted a
great and learned savant at court. Before that time, in the reign of
Henry the Seventh, there had been a namesake of Colin’s who had
fought bravely for the crown, and helped to subdue Lord Lovel’s
rising in Yorkshire. Claudia knew of these and of several more
worthy and later ancestors, for she had once visited his Elizabethan
country home, where his mother still lived, and he had, with
laughing comments, conducted her through the gallery of family
portraits, which showed, he said, that there had never been any
fatal beauty in the family. But she had been struck even then, as a
girl—she had only been seventeen at the time—with the indefinable
air of breeding and intellectual distinction which they all bore. There
was an unmistakable stamp on the faces of all the Patons, which
said as plainly as words, “Death before dishonour.” Colin had told her
the story of one youth, a gay Royalist with laughing eyes, who had
fallen from honour by parting, under pressure from the woman he
loved, with one of the King’s secrets. “But, like Judas,” said Colin,
“he went out and hanged, or rather shot, himself almost directly
afterwards. You, who feel so intensely the joy of life—look at his
laughing eyes!—will believe that he expiated his sin.”
At Gilbert’s home, too, there was a small picture-gallery—not very
large, for the Curreys had never had any artistic leanings, and had
only had their portraits painted to feed their own vanity and pomp—
but the Curreys were a different race. Worthy—yes, probably—but
heavy and coarse-featured, with none of the fineness and delicacy
that distinguished the Patons, and some of them obviously too full-
blooded, with the limited vision which embraces only the material
things of life.
The man who stood looking down upon her now was of different
type from either. He belonged to the virile new world; he had its
good qualities and its defects. Like Colin, he was a good companion
to be with, but he was so virile and so mettlesome that he
occasionally left her rather exhausted.
“Well?” he queried smilingly, not attempting to answer her
question.
“I was thinking.”
“I know you were. One can always see the thoughts flitting
through your eyes. I have often longed to know what you were
thinking about. I believe your thoughts are worth hearing. Won’t you
tell me this time?”
She found herself liking his voice, which had a slight American
inflection without being nasal.
“I was thinking how different the American man is from the
average Englishman, both in mind, temperament and physique.”
“We’re certainly beaten under the last head,” he replied, with a
frank laugh. “I am always admiring your Englishmen from the point
of view of good looks, though you know our men can be pretty fit,
as we’ve shown in your sports’ contests. But we’re not such good
lookers, sure. As for temperament”—he looked at her with a little
challenge in his grey-blue eyes—“that isn’t racial, you know; it’s
individual. I guess one of my countrymen may possess it as well as
an Englishman. And what do you mean by a temperament,
anyway?”
Claudia shook her head. She refused to be drawn. “Impossible to
define. Those who have it do not need a definition, and those who
have it not—will never find one. Didn’t someone once say: ‘Art is life
seen through a temperament’?”
“But I’m not an artist,” he replied quickly, “only a merchant, who
purveys works of art through the medium of a printing-press. Do you
think that only professed artists may possess a temperament?”
“Of course not. That would be too ridiculous. I daresay some of
the greatest artists are inarticulate.”
“I am glad to hear you say that, because I should have hated to
have you put me right out of court. Because,” he spoke slowly,
“lately I have begun to realize that a certain resurrection is going on
within me; that what I tried deliberately to kill is still alive, painfully
alive.”
She was aware that he was on the verge of a confidence, and she
only looked her interest. She liked him, and she felt she wanted to
know more about him; for never had they discussed their private
lives with one another. He was introducing a new element into their
friendship.
“I married before I was twenty-two, and last fall I became a
widower. I married early after deliberation and sober reflection. Isn’t
it curious that one can so often reflect more soberly when one is
twenty than when one is approaching forty, as I am now? I married,
my friends said, most suitably. I was not what you would call in love
with her. I had known her for years, and I was fond of her in a quiet,
unemotional way, which you people of temperament despise. I
married young to have my mind and energies free for my work of
restoring an old firm to its original activity and greatness. I realized
that if youth wants to toe the straight line, it must keep clear of
emotional complications. I saw other men taken off their work, their
senses flaying them into madness and folly, by the women they met.
I determined that I would marry and keep clear of attractive women.
I would settle down early into a family man, and if there were joys
that I knew not—well, the man who has been born blind doesn’t
know the glory of the sunshine. My wife was placid and quite
content with the small amount of leisure and attention I could give
her. All my best energies I gave to my work. Every American is born
ambitious; it’s in the very air he breathes, and with his first little
squalling breath he draws it in. I had rather a tough fight, but I won
out all right.... Now I am nearly forty I begin to wonder if I have
done the best with my life; I begin to see that perhaps those other
fellows who never got on are not to be pitied after all. I begin to feel
a hiatus in my life; I begin to see what life might be.”
As he looked at the beautiful vivid woman among the cushions of
the armchair, he recalled the quiet, orderly life he had led with the
one who had borne his name, the lack of anything approaching
exaltation or beauty in their relationship, the prosaic monotony of
their days, and he wondered if he had not been the greatest of
God’s fools. What would life be with such a woman as the one Who
now sat plaiting her fingers in her lap, her very finger-tips pulsating
with life? The magnetism of her womanhood reached him as he
stood, and made his breath come more quickly. They had so much in
common already, was it too wild and venturesome to hope that they
might have more?
“In short,” she said slowly, “you have sacrificed the best years of
your life to what you men call ‘the game.’ But you have succeeded.
Many men sacrifice everything and—fail. You may feel at odd
moments that you have missed something, but I expect you are
really quite satisfied. You know the proverb about the cake?”
“Yes, but did I choose the best kind of cake?”
She broke the spell by laughing. It sounded so odd. It reminded
her of the days when, as a child, she used to hover over the plate of
cakes anxiously seeking to make a good choice.
“That’s life,” she laughed. “If you take the chocolate one, you
always wish you had taken the jam-puff. And, after all,” a little
wearily, “what does it matter—chocolate or jam? Equally sweet,
perhaps, and equally unwholesome.”
He joined in her laugh and held out his hand. “I must go now. Let
me come again soon, will you? I enjoyed your charming luncheon-
party, but much more have I enjoyed this talk with you. Somehow I
always want to talk to you, and I have the reputation for being
rather a silent man. I wonder why you inspire me?”
Her hand was in his and she smiled mischievously and mockingly
as she said: “I suppose it’s because I talk so much. It makes you
feel that you must uphold the superior ability of your sex in all
things, even conversation.”
But he did not smile. His eyes were searching her face, noting the
soft, velvety texture of the skin—how he longed to press his lips on
her full, creamy throat even more than on her lips—the satiny gloss
of her luxurious hair, the long eyelashes which, as he stood above
her, swept her cheeks, the small, straight nose and delicate ears.
“You are a very sweet and fascinating woman,” he said suddenly,
“and I am sorry that we ever did anything so vulgar as to use your
portrait for a book cover.... Good-bye.”
For a few minutes after he had taken his departure Claudia sat
thinking about him. Unlike Frank Hamilton, he did not set her pulses
singing, and leave her inwardly shaken when he released her hand;
but, on the other hand, she found herself considering him more
seriously. She conjectured more about him; she found herself
wanting his opinion, just as she did Colin Paton’s. Colin! That
reminded her of the beginning of their conversation. Colin had
clearly shown that their friendship was to him but a small thing. She
found herself clenching her fingers into the palm of her hand as she
reflected on the secret he had kept from her. This man Littleton was
not in any way the equal of Colin Paton, either in brain or in
character; but he was evidently trying to tell her how much he
appreciated their acquaintanceship, trying to let her know that he
realized now what a big part a woman might play in his life. Pat was
quite, quite wrong. Colin was an unemotional fish; he even took
their friendship coldly.
“And I want love, life, warmth!” she cried to her empty drawing-
room. “I am tired of leading this deadly existence. I want someone
to love me, to tell me so, to make me feel that he loves me.”
She looked at the room through a blinding mist, so that the
delicate walls and the Louis Quinze furniture all swum in a haze, and
nothing stood out save the fact that the room, like her heart, was
empty, and there was no one to hold out two arms ready to enfold
her.
Then she strangled a sob in her throat, and the room became
once more the charming, orderly room it always was, filled with
sweet scented flowers and majestic palms.
“You’re a fool, Claudia, a fool! a fool! a fool!” she said through her
half-closed teeth. “You want things that you will never get, that
probably don’t exist except in your stupid imagination.”
Then she went quickly out of the room to her bedroom, where her
outdoor clothes were lying on the bed. She rang the bell for her
maid.
“Order the car for me, please. I am going to see Mrs. Iverson.
Give me that box of picture-puzzles I got for her.”
Fay always wanted her. She would go where she was wanted.

CHAPTER XV
WHY NOT?
Claudia asked the usual question of the nurse who met her in the
hall of the flat. It was now three weeks since Fay’s accident.
“Sir Richard said definitely to-day that everything has now been
tried,” said the nurse sadly, for both the day and the night nurse had
grown fond of their odd little patient. “I think they always knew it
was hopeless.... I fear she is growing suspicious. She cried a good
deal of last night, and only slept for a couple of hours. Nurse
Calderon said she thought she heard her whisper to herself in the
night: ‘Oh, God! I can’t! I can’t! Let me get better!’ Poor little thing!
It’s too horrible, and, of course, everything will—will get worse.”
Claudia, who had read up the progress of such cases in a medical
book she had found in Gilbert’s library, gave assent. She knew that
the end of such cases is the abject humiliation of human flesh where
so many of the functions of the body are paralysed. The account had
made her feel sick in the reading, and she shrank from the thought
of all that lay before the girl—she was little more—who lay in the
bedroom beyond.
Claudia opened the bedroom door full of misgivings, her heart
very heavy as the thought of Fay’s night vigil, so that she was
unprepared for the sight that met her gaze. The room always was a
bower of flowers, generally coloured ones, for Fay said bluntly that
white ones reminded her of a funeral; but this afternoon it presented
an unusually gay aspect. The apartment was almost gaudy, and at
first Claudia did not take in why it was so bright. Fay was propped
up among a nest of pillows, her tiny face, very little changed, hidden
under an enormous black hat with three great blue feathers floating
over it. The bed was strewn with hats, the chairs were littered with
them. Pink cardboard boxes of various sizes stood everywhere.
“Darling, you’ve come in the nick of time,” called out Fay excitedly.
“Isn’t this a duck of a hat? You see, I must have some new hats. I
shall be better soon now, and it’s no good getting up and finding
you’ve got nothing to put on your cocoanut. And Madame Rose has
got all her new models for the summer. This is French. You can see
that with half an eye, can’t you? I call it shick, don’t you? Something
like a hat.”
A dark-eyed Jewess, who had evidently brought the hats, was
standing at the foot of the bed, and broke in with:
“Straight from Parry, Miss Morris,” she said glibly, though it was
evident that it had been concocted in some cheap London
warehouse. “Very latest thing. Real style there. I thought of you as
soon as I saw it. It’s too good for anyone else, I said.”
“Ah! did you? Give me the hand-glass. I want to see how my dial
looks under it. Ugh! like an under-done muffin left out in the rain.
Give us over the rouge and the powder-puff. And the bunch of curls
out of the drawer. Where’s that eyebrow pencil I had this morning? I
rub the blessed stuff off on the pillow. There! that’s better, cocky.
Now I’ve got a bit of bloom. We’re not forty and in the cupboard yet,
thank the Lord! It saves a lot of trouble if you’ve got dark eyebrows.
Yours don’t rub off and get smeary, do they?”
“It’s curious,” smiled Claudia, removing one of the hats in order to
sit down, “that your eyebrows are so light when your hair is so
dark.”
Fay gave a whoop that showed her lungs were not affected.
“You dear holy innocent! Did you think my hair was really this
colour? Not much. The hair-dresser does it, and jolly expensive it is.
My hair, as a child, was a silly soppy sort of light shade, so I
improved on it. I’m much more effective with black hair. Makes a bit
of a contrast. Got the idea out of a story where a man was raving
over blue eyes and black hair. First of all, I tried red. But it’s so
difficult with hats and all the boys call you Ginger.”
She might have been discussing the colour of a parasol, so
impersonal and frank was her tone. Evidently it never occurred to
her that these were what is called in ladies’ papers, “secrets of the
toilet-table.”
Fay turned to the girl, who was adjusting the trimming on another
hat, equally large and covered with roses of a nightmare shade of
pink.
“You remember my hair when it was red, don’t you, Vera?” She
chuckled. “I remember you didn’t know me when I came into the
shop, and you was so polite”—she gave Claudia a wink—“that I
knew you hadn’t spotted me. I’d run up the devil of a bill, and
Madame Rose was giving me the frozen eye just then. I think I shall
keep to black now. It does suit me, doesn’t it?”
“Admirably,” returned her sister-in-law, controlling a desire to
laugh.
“I like your hair,” commented Fay; “there are sort of coloury bits in
it. I thought at first you must dye it, only Jack told me you didn’t,
and that it was like that when you were a kid. It’s real pretty.
Darling, try on this hat. I want to see it on someone else. There’s no
doubt it’s stylish. I hate the sort of hats nobody notices. When I pay
big money I like to get the goods.”
Claudia good-naturedly removed her own smart little toque of
white brocade and skunk, and placed the top-heavy confection upon
her head.
Fay’s face was a study in astonishment and dismay as she looked
at the other woman.
“Well, I’m blowed! It looks—oh! sort of funny—and”—she shook
her head—“Vera, are you sure it’s good style? All right, keep your
hair on, I didn’t say it wasn’t, only—— Crickey Bill, does it look like
that on me?”
The girl from the shop eyed Claudia with no great favour. Her
small, beady eyes looked sourly and enviously at the perfectly-cut,
black velvet gown and elegant skunk and ermine furs. She was cute
enough to realize that Claudia’s clothes were the “real thing” and
spelt not only money—her own wares were absurdly overpriced—but
taste. She was accustomed to serving “ladies” in the profession, who
familiarly called her “Vera, my dear,” and asked, and generally took
her advice, as well as swallowed her fulsome flattery.
“Take it off,” said Fay almost sharply. “I hate it now. It’s too large,
it’s too——” Then, with a sudden change to wistfulness, she added,
“but it’s you that makes it wrong. You’re good style, and I’m not. I’m
common, dead common. I don’t wonder you didn’t want me in the
family.”
“Fay, dear, don’t.” Claudia glanced at the sulky Vera, who was
packing up the hats. Apparently Fay had never heard of the
undesirability of washing dirty linen in public.
“You’re a lady. A blind man could see that. If you hadn’t been so
sweet I’d have hated you directly I saw you. I knew what you were
at once. Of course, Jack is a perfect gentleman, but that’s different
somehow, except”—vaguely—“I liked him a bit extra for it. He looks
different in his clothes to the other men, and yet those men spend a
lot of money too. I knew a man once, he owned a couple of halls in
the Midlands, and he told me he had fifty-two waistcoats, one for
every week of the year. I don’t suppose Jack’s got as many as that?”
She was adjusting a saucy matinée cap, a dainty affair of pink
ribbon and lace.
“I am sure he hasn’t.”
“Won’t you take no hat at all?” said the annoyed shop-girl,
breaking in rudely. “You might take this one with the pink roses. I’m
sure that’s quite enough.”
“No, no, I’ll wait till I can come to the shop. Here, my dear, here’s
a half a crown for your trouble. I’ll come in—soon.” She looked
quickly from the shop-girl to Claudia, a desperate question in her
blue eyes.
“That’s a much better arrangement,” returned Claudia cheerfully.
“We’ll go together, shall we?”
“Yes, yes,” cried Fay eagerly, clapping her hands. “But, I say,” as
the door closed behind the girl and her hat-boxes, “will you take me
to your hat shop where that came from?”
“With pleasure.”
“What; come here.” Fay beckoned her imperiously to her side. “Do
you mean you are not ashamed of me? I could keep my mouth dead
shut, you know. Do you mean that you’d let me wear the same sort
of hats as you, that you’ll try and make a lady of me?”
Claudia could not speak, she gently nodded.
“Well,” said Fay huskily, her eyes suspiciously moist, “you’re it all
right, that’s all I can say. I—you can touch me for anything you
want. You’ve only got to ask me. I say, hand me over that leather
case from the chest of drawers—yes, that’s the one.”
Wonderingly, Claudia obeyed, and handed her the case which was
a cheap leather imitation.
Fay opened the case with a key from under her pillow and
rummaged inside. Presently she produced a small box.
“There! I want to show you this. It’s for you. It’s quite straight;
you needn’t think I got it in any—any way you wouldn’t like. I
bought it off someone who was hard up.” “It” was a diamond and
ruby brooch, and quite a tasteful affair in the form of two hearts,
transfixed by an arrow.
“Oh! but Fay, I couldn’t——”
“Take it, I say, or I shall think you don’t mean what you said just
now. Two hearts, d’yer see—you and me! Quite romantic, isn’t it?
Put it on that lacy thing at your throat. Yes, it looks nice. No, you’re
not going to thank me. Just give me a kiss, that’s all.”
For a few moments the lips of the two met, so different in their
upbringing and views of life, but strangely brought together by the
hand of Fate.
“Now look at my joolery. Never seen it, have you? Well, it aint so
dusty, if I says it. I’ve always got them to shell out all right. After
all,” with a quaint little touch of vanity, “when you top the bill you’re
worth it, and I don’t believe in making yourself cheap or making
men meaner than they are. Not that I exactly like them for what
they give you, but it shows they do like you, because a man doesn’t
stump up easily.... There, that’s a stunning pendant, isn’t it? It cost
two hundred and fifty, because I went and chose it.”
Claudia was astounded at the value of the jewellery that reposed
in the shabby, unremarkable leather case. She saw that Fay loved
the things by the way she touched them. Some of them were
beautiful. But presently Fay gave a sigh and, selecting a large
diamond pendant which she put round her neck, over her
nightdress, she shut up the case. “Put the things back,” she said
queerly. “I—I——” Then, to Claudia’s dismay, she began to sob
rather pitifully like a frightened child. Claudia drew the little head to
her breast.
“Hush, dear, you mustn’t excite yourself. It’s bad for you. Nurse
will say it’s my fault, you know.”
“I’m not very old,” sobbed Fay, “I’m only twenty-two. Some people
live to be very old.”
Claudia tried to think of a laughing reply, but no words would
come. She could only rearrange the matinée cap and put her own
cool cheek against the one wet with tears.
“Fay, dear, to please me—you said you’d do anything for me—
don’t cry so. Are you—are you in pain?”
She wiped the tears away gently with her handkerchief, the rouge
from the cheeks coming off too.
Presently Fay grew a little calmer.
“Claudia, I want to ask you something because you are honest.”
Oh! how Claudia’s heart sank! She dreaded what the next words
would be, but as usual the unexpected came from Fay.
“Do you think this is a punishment for—for not being good? Nurse
has got a Bible, and I—just for fun—asked her to read me a bit. It
frightened me. I’m not what you call bad, am I?”
“No, Fay,” said Claudia steadily, determined that not all the religion
or moral teaching in the world should make her distress the doomed
woman. “No, Fay, don’t distress yourself. I don’t believe for an
instant this is a punishment.” She tried to speak simply, but the task
was difficult. Her own religion was a very vague one. She believed
that if there were a God, as so many Christians averred, a God who
was all-loving, understanding beyond finite conception, there could
never be any question of punishment such as Fay suggested. Fay’s
mind and morals were stunted, undeveloped. Since she had come in
contact with the queer people who were her fellow “pros,” Claudia
had come very clearly to recognize that the lives of such artistes,
especially those like Fay, who had been born practically on the
boards of a music-hall, were not subject to the ordinary judgments
of society. Theirs was a little world of its own, with its obligations, its
own ideas of right and wrong. To do another artiste out of a job, to
queer her turn, to refuse to put your hand in your pocket for a
deserving case, to crib another person’s business or her “fancy boy,”
those were unpardonable sins in Fay’s world. To have flitted from
lover to lover—in her case without any breaking of hearts or ugly
recriminations—was only a venial one.
Fay gave a huge relieved sigh. “If you say so, I won’t worry about
that any more. Of course, mind you, I ought to have kept straight.
Mother told me that when I was a kid. But I don’t know. Men always
liked me, you see, and I’m fond of them. Of course, I know you
wouldn’t do the things I’ve done.”
Claudia inwardly winced. That very morning she had had an
impassioned lover-like letter from Frank complaining that she never
came for the sittings now. “I know you have been a great deal with
your sister-in-law, but sometimes I fear you cannot care for me
when you can live without seeing me. To me, you are the whole
world.”
“I expect Jack and I are pretty poor tripe,” continued Fay calmly.
Then a new thought struck her. “I say, that night I fainted, I thought
I heard a nice voice in the hall, a man’s voice. It wasn’t the doctor,
because he’s got a down-in-your-boots voice, and it wasn’t none of
my pals. Was it someone, or did I fancy it?”
“I think it was probably a friend of mine, Colin Paton. He got the
specialist and nurse for you, and often inquires after you.”
“That’s jolly decent of him, because he doesn’t know me from
Adam.” She looked round her at the many vases crowded with
flowers. “But people have been nice to me, haven’t they? It shows
I’m liked, doesn’t it?” It was such harmless vanity that Claudia
smiled. “Is your friend a great swell, Sir Somebody or other?”
“Oh, dear, no.” Claudia found herself laughing at the idea of
anyone calling Colin Paton “a great swell.” She must remember to
tell him, he would enjoy the joke too. Then she stiffened a little. No,
she would not tell him anything. He left her out of his life. “He’s the
simplest and kindest of men, a friend one can always rely on.” Her
sense of fairness prompted her to say so much.
“He’s old, then?”
“No, about thirty-eight. Did my description sound like a
greybeard?”
“Yes, ‘kind’ sounds so old somehow. Of course, he’s gone on you.
He must be. Would he come and see me, do you think? Why,” with a
sudden flash of inspiration, “it must be the man Polly said was here
that night and treated her as if she was a duchess, and thanked her
for everything. Polly flopped immediate. She’s had a balmy look ever
since. Oh, yes, I don’t think! Is he handsome?”
“No, only nice looking.”
“Well, I should like him to have black, flashing eyes—don’t you
love black, flashing eyes—and dark curly hair, and long, white hands
like the man in the novel, ‘Did He Love Her.’ I’ll just have to listen to
his voice.... Must you go now? Oh, well, I suppose I mustn’t be
selfish. Jack will be in soon. It’s rough on Jack me being like this,
isn’t it? Only a log for a wife.... He’s better than I expected,
because”—with a canny wag of her head—“Jack didn’t marry me to
have me lying here, like this. Men like their women to be pretty lively
and ‘on the go,’ especially when they marry someone of my sort.
Poor old boy! I’m really fond of Jack, you know. He’s always treated
me decently. I hope I’ll get well or else—— All right, yes, of course, I
won’t worry. Come again to-morrow. Where are you going?”
“To my mother’s. She’s got a musical afternoon, and I must look
in. Several grand opera stars and a great pianist. It will be very
fireworky, I’m sure. Good-bye, dear.”
Fay kissed her hand gaily as Claudia smilingly withdrew.
In the hall she met Jack coming in.
“Hallo! Claud.” He heaved a deep sigh. “I say, this is breaking my
heart.”
“Don’t think about your heart, think about hers,” said Claudia,
putting her hand on his shoulder. He looked very dejected and some
of the youth had gone out of his face. The contented, well-fed
expression was flecked with something closely resembling
unhappiness. “She is not likely to live for many years, and let’s try
and make the best of it for her, Jacky boy.”
“It’s hell hearing her talk about her new songs and going to Paris
with me.... I shall blurt out the truth one day, sure as Fate. It’s lucky
I’ve got a stolid sort of look, but it breaks me up inside. I remember
talking to you once about thinking too much and rootling about for
meanings in life. Why should Fay have to die like this? She hasn’t
harmed anyone!”
Claudia shook her head and was silent. Many greater minds than
poor Jack’s had wrestled with that problem, and there had never
been, and never would be, any answer. With Jack, his belated
questioning was rather pathetic. He had never wanted to ask
questions, he had been content just to live, and now his happy-go-
lucky love for Fay had turned into tragedy.
As they stood there they could faintly hear the parrot in the
distance still calling, “Chuck it! Chuck it!” accompanied by a hoarse
chuckle that seemed to mock them with some uncanny knowledge.
The little hall was tidy now, but it meant that its volatile mistress
would never dash through it any more.
“I say, Claud,” said Jack, taking off his coat, “what’s come over
Gilbert? I went into court to-day—a fellow I know was interested in
an arbitration case, had money invested—and when we got there I
found Gilbert had been briefed. He started splendidly in that ‘listen
to me’ sort of manner, and then he got muddled. He couldn’t
remember the name of the firm he was speaking about, and he had
to ask his junior. Everybody was noticing it. Why, he used to have
such a ripping memory! What’s wrong with the works?”
Claudia was not so alarmed as she well might have been had she
known the symptoms of nerve breakdown.
“Perhaps he took the case up in a hurry, sometimes he has to do
that, you know.”
“No, he didn’t, because the fellow with me told me that he knew
he had been secured for the case a long time ago. I heard someone
say he was going to pieces.”
“He wants a holiday.... Mother will think I am never coming. Go in
and talk to Fay.”
He saw her into her car, and a few minutes later Claudia found
herself alighting on the red carpet outside her old home. The sounds
of a violin played by a master hand reached her as she entered. The
Rivingtons were just going, Mrs. Rivington very shrill and chatty, and
the General rather tottery and deaf.
“I say,” said Mrs. Rivington, with a glint of malice in her eye, “is it
true your friend Frank Hamilton is going to marry Mrs. Jacobs? Good
thing for him, I should say. She’s just rolling in money, almost
indecent, and anyone can see she’s madly in love with him. It’s all
very well to talk art,” sneeringly, “but it usually spells money, doesn’t
it? Artists are just like the rest of us, only they pretend a bit more.
He’s always with her, so I suppose the engagement will be
announced soon.”
Claudia attributed the remarks to ill-nature on Mrs. Rivington’s
part, for her chief occupation in life was planting arrows as often as
she could in the weak spots in her friends’ armour. Claudia could
afford to smile serenely in reply. Did she not know whom Frank
loved? A woman rather enjoys a clandestine love-affair, and Claudia
hugged to herself her closer knowledge of Frank’s inner life. She
knew she was the core of it.
“Mr. Hamilton’s in there now, talking to the Duchess of Roxford,”
continued Mrs. Rivington. “Ridiculous how artists are run after, isn’t
it? I don’t suppose he was anyone in particular. Artists never are.
Some people find that interesting, but I must say, personally, I
prefer good breeding. So unmistakable. Good-bye. It’s too dreadful
about The Girlie Girl, but I was right, after all, wasn’t I?”
Claudia stood quietly in the doorway until the violinist, the great
Ysaye, had finished playing. There were many well-known people
present, great names in the social and artistic firmaments, for Circe
had always held a little court all her life, and she had cleverly
managed to pursue her uneven way without offending any of the
powerful social leaders, who, though they always remembered her
trespasses against her, generously spoke with more or less
indulgence of them. She was hated by a few, like Lady Currey, but
they did not count for very much. Circe had never been actively
malicious, and she had always been too immersed in her own affairs
to find time to be inquisitive about other people’s, hence she had
acquired a certain reputation for fair dealing and generosity of
character not altogether deserved. Now she very seldom
entertained, but when she did so, she did it superlatively well, and
many artists she had encouraged in their young and aspiring days
were glad to do her honour.
The music stopped and she found Frank at her side.
“At last! I have been waiting for you all the afternoon. I was afraid
you were not coming. Claudia, this cannot go on. You are driving me
mad. It is deliberate? Have you all the time just been playing with
me?”
“Hush! don’t be so indiscreet.” She smiled, for Mrs. Rivington’s
words returned to her mind. Frank Hamilton attracted by Mrs.
Jacob’s money-bags! “I’ll talk to you later. You shall get me some
tea. I must go over and speak to mother.”
She threaded her way, with handshakes and smiles, to where
Circe, in a most exquisite frock, sat in a shaded corner, among a lot
of scented cushions. She was talking with more animation than usual
to a man whose back was towards Claudia. With her quick eye for
beauty, she noticed that he had a particularly well-shaped head,
which was finely set on his shoulders. Circe was talking in French to
him.
“Eh bien, mon cher, Claudia est très belle, et elle est—”
Circe caught sight of her, and stopped short. Had it not been
almost impossible, Claudia would have thought that her mother
looked distinctly embarrassed and taken aback. Then the well-known
sweet smile drifted over her still beautiful mouth, and the
momentary impression vanished.
“Claudia, we were just talking of you. You are late, child. Let me
introduce to you an old friend, Mr. Mavrocopoulos.”
The man rose and bowed with unusual grace, and Claudia saw a
very well-preserved man of about fifty-five, with black hair flecked
with grey, and remarkably fine dark eyes. She returned his evident
look of interest, and again she received a peculiar impression as of
something that was vaguely familiar and yet somewhat dreamlike.
She was aware that Circe was watching them.
“Have I not met you before?” inquired Claudia. “Your face seems
familiar to me, somehow.”
Something flashed into his eyes, and his lips smiled as he turned
to Circe.
“No, Claudia, I don’t think you can remember Mr. Mavrocopoulos.
He has not been in England for many years.”
“But I saw you when you were a child of three,” said the man. “I
remember you well, very well. I do not pretend that I should have
known you as that child, but I remember you well.”
Claudia knew his name as that of a famous and very wealthy
Greek family, and she recalled a rumour that had once linked it with
her mother’s. Had they found happiness together? Were there
golden memories between them? She wondered curiously how a
man and woman felt in such a case, who, after the lapse of many
years, met again. Did yesterday seem as to-day? Was memory sharp
or dulled by time, did they remember the high-water-mark of their
passion, or the moment when they had said good-bye? Were they
glad to meet again? If she and Frank met after many years, would
they——? Then suddenly she heard Fay’s voice saying confidently: “I
know you wouldn’t do the things I’ve done.” But Circe had done
them, too, and she had not had the excuse poor Fay could bring
forward.
There were no signs of regret on her mother’s face. She never
spoke as one who finds any bitterness in the dregs of such a past.
Indeed, she always spoke as one who felt that she had fulfilled her
destiny, who has eaten stolen fruit joyously, without a scruple,
without a fear. Her mother’s contempt was for women who looked
longingly over the hedge and were afraid to jump.
With a few more words Claudia left the two together.
Circe’s slanting eyes, carefully made up, but in the shaded light
still siren-like and magnetic, looked for some seconds into the eyes
of the man beside her.
“She is like you, Demetrius, and she has always been my
favourite,” she murmured.
His only answer was to take her hand in his, and raise it to his
lips.
“I return to Rome next week, but I take back with me a living
picture, the incarnation of a dream.”
Claudia was sipping the cup of tea that Frank had procured for
her, when she bethought herself that she had not yet seen Patricia.
“Have you seen Pat? It is not humanly possible that she has
tucked herself in a corner!”
His eyes were hungrily devouring her face, and lingering on her
lips, so that she had the pleasant sensation of a secret caress. Mrs.
Jacobs! How ridiculous!
“I saw her disappear half an hour ago in a conspirator-like manner
with Mr. Colin Paton, into that room over there.”
He pointed to a closed door, which was the door of the library.
“Nonsense. What have they got to conspire about?”
There was a little frown between her brows. Colin was her friend.
“Why do men and women usually conspire to be alone together?”
Without answering, Claudia crossed the hall, and abruptly turned
the handle of the library-door.
Seated close together, talking very earnestly, Pat more excited
than she had ever seen her, were the two whom Frank had seen
disappear half an hour before. As a matter of fact, it had only been
ten minutes, but Frank had always had his doubts of Colin’s
friendship.
“ ... bushels of apples and immense quantities of ...” Pat was
saying, when her sister came in. “Oh! Claudia, you have come. We’d
almost given you up.”
In an utterly different style from her own, Patricia was looking
most attractive that afternoon. She had on a soft white charmeuse
gown, which showed the long lines of her figure, and clung around
her in a manner calculated to send her admirers crazy. The cool
nonchalant look which she usually wore had given place to
something more intense, more human. Something seemed to have
aroused her from her virginal slumber, and is not that brightness in
the eyes, that flush on the cheek, generally aroused by a male?
Claudia took all this in at a glance, and it was not till afterwards that
she had time to reflect on the odd subject-matter of their earnest
conversation.
“I wondered where you were,” said Claudia, rather frigidly. “How
do you do, Colin? I think mother wants you, Pat.” It was a fib, but
she had to explain her entrance.
Then she turned with a sweet but cold smile to Colin Paton, who
had quietly risen.
“I hear you have written a great book and are going to become
famous. Congratulations! I must buy a copy as soon as it comes
out.... Frank, I want some more tea. I’m so thirsty.”
Pachmann was playing as they made their way back to the tea-
room, his fairy-like fingers lightly caressing the keys into exquisite
joyousness.
“I want you to come to the studio to dinner next Monday,” said
Frank eagerly. “You always said you’d like to meet Henry Bridgeman
and his wife if I could arrange it?” Claudia was a great admirer of
Bridgeman’s etchings. “Well, they are coming to dinner at the studio
on Monday. Will you come too?”
“Of course, I shall be delighted,” returned Claudia, not even
troubling to think of her engagements. “I shall love it. And”—with a
hard laugh—“I’ll come for a sitting to-morrow if you like, before I go
to Fay.... Dear, you mustn’t say such things here. It’s compromising.”
A loud chord on the piano, immediately followed by the sound of a
man’s voice, made her raise a warning finger. “Hush!”
The words came clearly enough to both of them as they stood
together.
“Ah! fill the Cup, what boots it to repeat,
How Time is slipping underneath our feet:
Better be jocund with the fruitful grape
Than sadden after none, or bitter fruit.”

It was Liza Lehmann’s setting, and the accompaniment thundered


and rumbled, and then softened down to a plaintive, appealing
melody. It might have been the voice of Circe herself, beckoning,
alluring, promising....

“Ah! love, could you and I with Fate conspire


To grasp the sorry scheme of things entire
Would we....”

After all, why had she so many scruples? How did she come to be
possessed of them? Why did she hesitate to grasp her happiness?
She looked up and found Colin Paton’s eyes fixed upon her, and
they wore an expression she did not know.
Then she heard Frank’s voice murmuring in her ear. “Claudia, if
you only knew how much I love you. If you would only trust yourself
to me. Why are you afraid?”
“I don’t know,” she said truthfully, “I don’t know.”
She gave him a particularly tender smile, out of sheer feminine
perverseness, impelled by something that rankled and festered
within her. Colin Paton should be made to understand that there was
at least one man who was a real friend to her, yes, and might be
more.

“Turn down an empty Glass....”


Why not?

CHAPTER XVI
NATURE’S FAULT

Claudia was leisurely dressing for the dinner à quatre at Frank’s


studio, leisurely, because there was something in the warm May air,
stealing in through the windows, that made her dawdle and dream.
She and Pat had motored out into the country that morning, and
lunched at a quaint old inn covered with wistaria, just outside
Penshurst, and the spell of the country, with its riot of scent and
song, still possessed her. She thought of the hedges, with their
tender greens; the young grass studded with gold and silver, for the
buttercups and daisies were gaily blooming; the lilac in the cottage-
gardens, just bursting into exquisite flower; the primroses with their
pale beauty, nestling at the roots of the trees; the fruit blossom
making a poem in delicate pinks and whites. She looked at the bowl
of wild hyacinths she and Pat had gathered as excitedly as a couple
of Cockney children, and she wished that she could have stayed in
fairyland a little longer. She had been so happy for a few hours, for
she loved the country. She had put away all the problems that beset
her, and she had let the sweet perfection of Nature soothe her into
something closely resembling peace. She had given herself up to its
healing, and she was still between it and noisy nerve-racking London
as she donned her clothes. In accordance with her mood, she had
chosen to wear a simple, almost girlish dress of faint pinks, that
reminded her of the orchards they had passed through, and, as a
finishing touch to remind her of their excursion, she pinned some
primroses on her corsage. Their delicate perfume was like fresh
honey.
Her maid noticed that she looked very young that night, with the
dreams in her eyes and on her lips, even younger than her twenty-
three years. Usually she looked much older, for her self-possessed
manner, inherited from her mother, her dignified carriage and air of
savoir faire might have belonged to a woman of twenty-eight. To-
night she almost had the illusion that she was still an unmarried girl,
with The Great Choice before her. The soft, warm air seemed to
breathe love, to say, “Take your fill of its sweetness, your life is still
to make.” The impassioned song of the birds, the riot and colour, the
bursting life in bud and blossom, what did it all say, but:

“Come, all lovers, to the feasting,


Where the wine of life is yeasting,
Soul of human, brute or flower,
This your purest, fullest hour
Drink your fill of Love’s own brew.”

Even Rhoda Carnegie’s cynical words the previous evening at the


Prime Minister’s dinner-party seemed part of the day. “Is love to be
confined within the small circlet of a wedding-ring? Why, it would be
like trying to pour the sea into a thimble.” After all, most intelligent
people nowadays scoffed at the wedding-service, with its “forevers”
and “till death.” Those ideas had all been swept away.
As she rearranged the wild hyacinths for the mere pleasure of
touching them, she asked herself if there still lingered any belief in
those “forevers.” Honestly, no. She did realize that love is too big a
thing to be confined within a wedding-ring. It was not that kind of
scruple that held her back. Love, as she had once said before her
marriage, was the only convention she owned. She recalled the
words of James Hinton. “Love, and do as you please.” Many people
had taken this as their text for lax morality, but they had not
understood him rightly. It was not an easy saying, but a hard one.
Love! How often did one love in a lifetime? She had thought she
loved Gilbert, and she really had at the time. But his neglect and
coldness had killed her love. Could a great love be killed? “Many
waters cannot quench love——” was that not merely the high
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and


personal growth!

ebookmass.com

You might also like