Get ebook downloads in full at ebookname.

com

Metaheuristics for Machine Learning Algorithms and

Applications 1st Edition Kanak Kalita

https://ebookname.com/product/metaheuristics-for-machine-
learning-algorithms-and-applications-1st-edition-kanak-
kalita/

OR CLICK BUTTON

DOWNLOAD EBOOK

Explore and download more ebook at https://ebookname.com

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Handbook of Research on Machine Learning Applications

and Trends Algorithms Methods and Techniques 1st
Edition Emilio Soria Olivas

https://ebookname.com/product/handbook-of-research-on-machine-
learning-applications-and-trends-algorithms-methods-and-
techniques-1st-edition-emilio-soria-olivas/

Computer Vision Principles Algorithms Applications

Learning 5th Edition E.R. Davies

https://ebookname.com/product/computer-vision-principles-
algorithms-applications-learning-5th-edition-e-r-davies/

Introduction to Statistical Relational Learning

Adaptive Computation and Machine Learning Adaptive
Computation and Machine Learning Series Lise Getoor

https://ebookname.com/product/introduction-to-statistical-
relational-learning-adaptive-computation-and-machine-learning-
adaptive-computation-and-machine-learning-series-lise-getoor/

New Regionalism and Asylum Seekers 1st Edition Susan

Kneebone

https://ebookname.com/product/new-regionalism-and-asylum-
seekers-1st-edition-susan-kneebone/
Democratic Theorizing from the Margins 1st Edition
Marla Brettschneider

https://ebookname.com/product/democratic-theorizing-from-the-
margins-1st-edition-marla-brettschneider/

Martin Rauch Refined Earth Construction Design with

Rammed Earth Detail Special 1st Edition Marko Sauer

https://ebookname.com/product/martin-rauch-refined-earth-
construction-design-with-rammed-earth-detail-special-1st-edition-
marko-sauer/

Negotiating Buck Naked Doukhobor Public Policy and

Conflict Resolution 1st Edition Gregory J. Cran

https://ebookname.com/product/negotiating-buck-naked-doukhobor-
public-policy-and-conflict-resolution-1st-edition-gregory-j-cran/

Rapid System Prototyping with FPGAs Accelerating the

Design Process Embedded Technology 1st Edition Cofer

https://ebookname.com/product/rapid-system-prototyping-with-
fpgas-accelerating-the-design-process-embedded-technology-1st-
edition-cofer/

Syntactic Argumentation and the Structure of English

David M. Perlmutter

https://ebookname.com/product/syntactic-argumentation-and-the-
structure-of-english-david-m-perlmutter/
Che Guevara Talks to Young People 2nd Edition Ernesto
Che Guevara

https://ebookname.com/product/che-guevara-talks-to-young-
people-2nd-edition-ernesto-che-guevara/
本书版权归John Wiley & Sons Inc.所有
 Metaheuristics for
Machine Learning

本书版权归John Wiley & Sons Inc.所有

 Scrivener Publishing
100 Cummings Center, Suite 541J
Beverly, MA 01915-6106

Artificial Intelligence and Soft Computing for Industrial Transformation

Series Editor: Dr S. Balamurugan ([email protected])

The book series is aimed to provide comprehensive handbooks and reference books for the
benefit of scientists, research scholars, students and industry professional working towards
next generation industrial transformation. .

Publishers at Scrivener
Martin Scrivener ([email protected])
Phillip Carmical ([email protected])

本书版权归John Wiley & Sons Inc.所有

 Metaheuristics for
Machine Learning

Algorithms and Applications

Edited by
Kanak Kalita
Vel Tech University, Avadi, India

Narayanan Ganesh
Vellore Institute of Technology, Chennai, India
and
S. Balamurugan
Intelligent Research Consultancy Services, Coimbatore, Tamilnadu, India

本书版权归John Wiley & Sons Inc.所有

 This edition first published 2024 by John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
and Scrivener Publishing LLC, 100 Cummings Center, Suite 541J, Beverly, MA 01915, USA
© 2024 Scrivener Publishing LLC
For more information about Scrivener publications please visit www.scrivenerpublishing.com.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or other-
wise, except as permitted by law. Advice on how to obtain permission to reuse material from this title
is available at http://www.wiley.com/go/permissions.

Wiley Global Headquarters

111 River Street, Hoboken, NJ 07030, USA

For details of our global editorial offices, customer services, and more information about Wiley prod-
ucts visit us at www.wiley.com.

Limit of Liability/Disclaimer of Warranty

While the publisher and authors have used their best efforts in preparing this work, they make no rep­
resentations or warranties with respect to the accuracy or completeness of the contents of this work and
specifically disclaim all warranties, including without limitation any implied warranties of merchant-­
ability or fitness for a particular purpose. No warranty may be created or extended by sales representa­
tives, written sales materials, or promotional statements for this work. The fact that an organization,
website, or product is referred to in this work as a citation and/or potential source of further informa­
tion does not mean that the publisher and authors endorse the information or services the organiza­
tion, website, or product may provide or recommendations it may make. This work is sold with the
understanding that the publisher is not engaged in rendering professional services. The advice and
strategies contained herein may not be suitable for your situation. You should consult with a specialist
where appropriate. Neither the publisher nor authors shall be liable for any loss of profit or any other
commercial damages, including but not limited to special, incidental, consequential, or other damages.
Further, readers should be aware that websites listed in this work may have changed or disappeared
between when this work was written and when it is read.

Library of Congress Cataloging-in-Publication Data

ISBN 978-1-394-23392-2

Cover image: Pixabay.Com

Cover design by Russell Richardson

Set in size of 11pt and Minion Pro by Manila Typesetting Company, Makati, Philippines

Printed in the USA

10 9 8 7 6 5 4 3 2 1

本书版权归John Wiley & Sons Inc.所有

 Contents

Foreword xv
Preface xvii
1 Metaheuristic Algorithms and Their Applications in Different
Fields: A Comprehensive Review 1
Abrar Yaqoob, Navneet Kumar Verma and Rabia Musheer Aziz
1.1 Introduction 2
1.2 Types of Metaheuristic Algorithms 3
1.2.1 Genetic Algorithms 3
1.2.2 Simulated Annealing 5
1.2.3 Particle Swarm Optimization 7
1.2.4 Ant Colony Optimization 8
1.2.5 Tabu Search 10
1.2.6 Differential Evolution 11
1.2.7 Harmony Search 12
1.2.8 Artificial Bee Colony 13
1.2.9 Firefly Algorithm 14
1.2.10 Gray Wolf Optimizer 14
1.2.11 Imperialist Competitive Algorithm 14
1.2.12 Bat Algorithm 15
1.2.13 Cuckoo Search 15
1.2.14 Flower Pollination Algorithm 16
1.2.15 Krill Herd Algorithm 17
1.2.16 Whale Optimization Algorithm 17
1.2.17 Glowworm Swarm Optimization 18
1.2.18 Cat Swarm Optimization 18
1.2.19 Grasshopper Optimization Algorithm 19
1.2.20 Moth–Flame Optimization 19
1.3 Application of Metaheuristic Algorithms 20
1.4 Future Direction 25

本书版权归John Wiley & Sons Inc.所有

 vi Contents

1.5 Conclusion 26
References 26
2 A Comprehensive Review of Metaheuristics
for Hyperparameter Optimization in Machine Learning 37
Ramachandran Narayanan and Narayanan Ganesh
2.1 Introduction 38
2.1.1 Background and Motivation 38
2.1.2 Scope of the Review 38
2.1.3 Organization of the Paper 39
2.2 Fundamentals of Hyperparameter Optimization 39
2.2.1 Introduction to Hyperparameters 40
2.2.2 Importance of Hyperparameter Optimization 40
2.2.3 Performance Metrics for Hyperparameter
Optimization 41
2.2.4 Challenges in Hyperparameter Optimization 41
2.3 Overview of Metaheuristic Optimization Techniques 42
2.3.1 Definition and Characteristics of Metaheuristics 42
2.3.2 Classification of Metaheuristic Techniques 42
2.4 Population-Based Metaheuristic Techniques 43
2.4.1 Genetic Algorithms 44
2.4.2 Particle Swarm Optimization 44
2.4.3 Differential Evolution 44
2.4.4 Ant Colony Optimization 45
2.4.5 Biogeography-Based Optimization 45
2.4.6 Cuckoo Search 45
2.4.7 Gray Wolf Optimizer 45
2.4.8 Whale Optimization Algorithm 46
2.4.9 Recent Developments in Population-Based
Metaheuristics 46
2.5 Single Solution-Based Metaheuristic Techniques 47
2.5.1 Simulated Annealing 47
2.5.2 Tabu Search 47
2.5.3 Harmony Search 48
2.5.4 Bat Algorithm 48
2.5.5 Recent Developments in Single Solution-Based
Metaheuristics 48
2.6 Hybrid Metaheuristic Techniques 49
2.6.1 Genetic Algorithm and Particle Swarm
Optimization Hybrid 49
2.6.2 Genetic Algorithm and Simulated Annealing Hybrid 49

本书版权归John Wiley & Sons Inc.所有

 Contents vii

2.6.3 Tabu Search and Particle Swarm Optimization Hybrid 50

2.6.4 Recent Developments in Hybrid Metaheuristics 50
2.7 Metaheuristics in Bayesian Optimization 50
2.7.1 Background of Bayesian Optimization 51
2.7.2 Gaussian Process Regression 51
2.7.3 Acquisition Functions 51
2.7.4 Recent Developments in Metaheuristic-Based
Bayesian Optimization 52
2.8 Metaheuristics in Neural Architecture Search 53
2.8.1 Introduction to Neural Architecture Search 53
2.8.2 Applications of Metaheuristics in Neural Architecture
Search 53
2.8.3 Recent Developments in Metaheuristic-Based Neural
Architecture Search 54
2.9 Comparison of Metaheuristic Techniques
for Hyperparameter Optimization 55
2.9.1 Criteria for Comparison 55
2.9.2 Comparative Analysis of Metaheuristic Techniques 55
2.9.3 Performance Evaluation of Metaheuristic Techniques 58
2.10 Applications of Metaheuristics in Machine Learning 61
2.10.1 Supervised Learning 61
2.10.2 Unsupervised Learning 62
2.10.3 Reinforcement Learning 62
2.10.4 Deep Learning 63
2.11 Future Directions and Open Challenges 63
2.11.1 Opportunities for Improvement in Metaheuristics 64
2.11.2 Adapting Metaheuristics to New Machine Learning
Paradigms 64
2.11.3 Addressing the Computational Complexity
and Scalability 65
2.12 Conclusion 65
References 66
3 A Survey of Computer-Aided Diagnosis Systems for Breast
Cancer Detection 73
Charu Anant Rajput, Leninisha Shanmugam and Parkavi K.
3.1 Introduction 73
3.2 Procedure for Research Survey 77
3.3 Imaging Modalities and Their Datasets 77
3.3.1 Histopathological WSI 77
3.3.2 Digital Mammography 78

本书版权归John Wiley & Sons Inc.所有

 viii Contents

3.3.3 Ultrasound 79
3.3.4 Magnetic Resonance Imaging 80
3.3.5 Infrared Breast Thermal Images 80
3.4 Research Survey 83
3.4.1 Histopathological WSI 83
3.4.1.1 Machine Learning-Based Histopathological
WSI 83
3.4.1.2 Deep Learning-Based Histopathological WSI 83
3.4.2 Digital Mammogram 84
3.4.2.1 Machine Learning-Based Digital
Mammogram 84
3.4.2.2 Deep Learning-Based Digital Mammogram 85
3.4.3 Ultrasound 86
3.4.3.1 Machine Learning-Based Ultrasound 86
3.4.3.2 Deep Learning-Based Ultrasound 86
3.4.4 MRI-Based 87
3.4.4.1 Machine Learning-Based MRI Analysis 87
3.4.4.2 Deep Learning-Based MRI Analysis 88
3.4.5 Thermography-Based 89
3.4.5.1 Machine Learning-Based Thermography
Analysis 89
3.4.5.2 Deep Learning-Based Thermography Analysis 90
3.5 Conclusion 90
3.6 Acknowledgment 91
References 92
4 Enhancing Feature Selection Through Metaheuristic
Hybrid Cuckoo Search and Harris Hawks Optimization
for Cancer Classification 95
Abrar Yaqoob, Navneet Kumar Verma, Rabia Musheer Aziz
and Akash Saxena
4.1 Introduction 96
4.2 Related Work 99
4.3 Proposed Methodology 104
4.3.1 Cuckoo Search Algorithm 104
4.3.2 Harris Hawks Algorithm 106
4.3.3 The Proposed Hybrid Algorithm 110
4.3.4 Classifiers Used 113
4.3.4.1 KNN Classifier 113
4.3.4.2 SVM Classifier 113
4.3.4.3 NB Classifier 113
4.3.4.4 mRMR 114

本书版权归John Wiley & Sons Inc.所有

 Contents ix

4.4 Experimental Setup 115

4.4.1 The Compared Algorithms 117
4.4.2 Parameter Setting 118
4.5 Results and Discussion 119
4.5.1 Experimental Results of the Proposed Algorithm
With the SVM Classifier 119
4.5.2 Experimental Results of the Proposed Algorithm
With the KNN Classifier 122
4.5.3 Experimental Results of the Proposed Algorithm
With the NB Classifier 125
4.5.4 Comparison of the Proposed Algorithm Compared
to Other Recently Published and Popular Algorithms
for Cancer Classification 128
4.6 Conclusion 130
References 130
5 Anomaly Identification in Surveillance Video Using
Regressive Bidirectional LSTM with Hyperparameter
Optimization 135
Rajendran Shankar and Narayanan Ganesh
5.1 Introduction 136
5.2 Literature Survey 137
5.3 Proposed Methodology 138
5.3.1 Dataset 139
5.3.2 Preprocessing Using Normalization 140
5.3.3 Feature Extraction Using Video Swin Transformer 140
5.3.4 Anomaly Detection 141
5.3.4.1 Regressive Bidirectional LSTM 141
5.3.4.2 Hyperparameter Optimization 142
5.4 Result and Discussion 143
5.5 Conclusion 146
References 147
6 Ensemble Machine Learning-Based Botnet Attack Detection
for IoT Applications 149
Suchithra M.
6.1 Introduction 150
6.2 Literature Survey 151
6.3 Proposed System 152
6.3.1 Dataset 152

本书版权归John Wiley & Sons Inc.所有

 x Contents

6.3.2 Data Processing Using Normalization 154

6.3.3 Feature Extraction Using Principal Component
Analysis 154
6.3.4 Random Tree-Adaptive Artificial Neural Network 155
6.4 Results and Discussion 156
6.4.1 Accuracy 156
6.4.2 Precision 158
6.4.3 Recall 158
6.4.4 F-Measure 158
6.5 Conclusion 160
References 160
7 Machine Learning-Based Intrusion Detection System
with Tuned Spider Monkey Optimization for Wireless
Sensor Networks 163
Ilavendhan Anandaraj and Kaviarasan Ramu
7.1 Introduction 164
7.1.1 Intrusion Detection System 164
7.1.2 Machine Learning 165
7.2 Literature Review 166
7.3 Proposed Methodology 168
7.3.1 Problem Statement 168
7.3.2 Methodology 168
7.3.3 Data Collection 168
7.3.4 Data Preprocessing 170
7.3.5 Support Vector Machine 170
7.3.6 Tuned Spider Monkey Optimization 171
7.4 Result and Discussion 173
7.4.1 Accuracy 173
7.4.2 Precision 174
7.4.3 Recall 174
7.4.4 F1 Score 175
7.5 Conclusion 177
References 177
8 Security Enhancement in IoMT‑Assisted Smart Healthcare
System Using the Machine Learning Approach 179
Jayalakshmi Sambandan, Bharanidharan Gurumurthy
and Syed Jamalullah R.
8.1 Introduction 180
8.2 Literature Review 182

本书版权归John Wiley & Sons Inc.所有

 Contents xi

8.3 Proposed Methodology 184

8.3.1 Data Collection 185
8.3.2 Data Preprocessing 185
8.3.3 Support Vector Machine 185
8.3.4 Multilayer Particle Swarm Optimization 187
8.3.5 Performance Evaluation 189
8.3.5.1 Accuracy 190
8.3.5.2 Precision 190
8.3.5.3 Sensitivity 191
8.3.5.4 Specificity 191
8.3.5.5 Security 192
8.4 Conclusion 192
References 193
9 Building Sustainable Communication: A Game-Theoretic
Approach in 5G and 6G Cellular Networks 195
Puppala Ramya, Tulasidhar Mulakaluri, Chebrolu Yasmina,
Pandi Bindu Madhavi and Vijay Guru Balaji K. S.
9.1 Introduction 196
9.2 Related Works 196
9.3 Methodology 197
9.3.1 Nash Equilibrium 202
9.3.2 Unique Nash Equilibrium 203
9.3.3 Total Unique Equilibrium 204
9.4 Result 207
9.5 Conclusion 211
References 212
10 Autonomous Vehicle Optimization: Striking a Balance
Between Cost-Effectiveness and Sustainability 215
Vamsidhar Talasila, Sagi Venkata Lakshmi Narasimharaju,
Neeli Veda Vyshnavi, Saketh Naga Sreenivas Kondaveeti,
Garimella Surya Siva Teja and Kiran Kumar Kaveti
10.1 Introduction 216
10.2 Methods 219
10.2.1 Competition and Sustainable Supply Chains
at Odds 220
10.2.2 Industry’s Effect Application of the 4.0 Idea
to Business Operations 221
10.3 Results 224
10.4 Conclusions 231
References 231

本书版权归John Wiley & Sons Inc.所有

 xii Contents

11 Adapting Underground Parking for the Future:

Sustainability and Shared Autonomous Vehicles 235
Vamsidhar Talasila, Madala Pavan Pranav Sai,
Gade Sri Raja Gopala Reddy, Vempati Pavan Kashyap,
Gunda Karthik and K. V. Panduranga Rao
11.1 Introduction 236
11.2 Related Works 237
11.3 Methodology 238
11.3.1 Framework for Research 238
11.3.2 Area of Research 238
11.3.3 Collection of Data 239
11.3.4 Analysis of Clusters 239
11.3.5 DSR Model 239
11.3.6 Model with Multiple Objectives 243
11.3.7 Statistical Data 245
11.4 Analysis 245
11.4.1 Feature Types 245
11.4.2 Timing of Renewal 247
11.4.3 Replacement of Function 248
11.5 Conclusion 250
References 251
12 Big Data Analytics for a Sustainable Competitive Edge:
An Impact Assessment 253
Rajyalakshmi K., Padma A., Varalakshmi M., Suhasini A.
and Chiranjeevi P.
12.1 Introduction 254
12.2 Related Works 255
12.3 Hypothesis and Research Model 255
12.3.1 Theoretical Model 255
12.3.1.1 Hypothesis 1 (H1) 256
12.3.1.2 BDAC as a Facilitator for IC (H2) 257
12.3.1.3 The Contribution of BDAC and IC
to Raising SCA (H3) and (H4) 257
12.3.1.4 Sustainable Competitive Advantage 257
12.3.2 Proposed Method 258
12.3.2.1 Layout Assessment 258
12.3.2.2 Analysis Method and Data Collection 258

本书版权归John Wiley & Sons Inc.所有

 Contents xiii

12.4 Results 259

12.4.1 Evaluation of Validity and Reliability 259
12.4.2 Structural Model of Coefficient
of Determination (R2) 263
12.4.3 Discussion 264
12.5 Conclusion 264
References 265
13 Sustainability and Technological Innovation in Organizations:
The Mediating Role of Green Practices 267
Rajyalakshmi K., Rajkumar G. V. S., Sulochana B.,
Rama Devi V. N. and Padma A.
13.1 Introduction 268
13.2 Related Work 269
13.2.1 Hypothesis Development 271
13.3 Methodology 271
13.3.1 Sample and Population 272
13.3.2 Techniques 272
13.3.3 Control 273
13.3.4 Data Analysis 274
13.3.5 Verifying Factor Evaluation 274
13.3.6 Standard Technique Bias 275
13.3.7 Correlation 277
13.3.8 Robustness Checks 278
13.3.9 Interaction Analysis 278
13.4 Discussion 279
13.5 Conclusions 281
References 281
14 Optimal Cell Planning in Two Tier Heterogeneous Network
through Meta-Heuristic Algorithms 285
Sanjoy Debnath, Amit Baran Dey and Wasim Arif
14.1 Introduction 285
14.2 System Model and Formulation of the Problem 288
14.3 Result and Discussion 295
14.4 Conclusion 298
References 298

本书版权归John Wiley & Sons Inc.所有

 xiv Contents

15 Soil Aggregate Stability Prediction Using a Hybrid Machine

Learning Algorithm 301
M. Balamurugan
15.1 Introduction 302
15.2 Related Works 303
15.3 Proposed Methodology 303
15.3.1 Soil Samples and Characteristics 304
15.3.2 Analyzing Soil Samples 304
15.3.3 Hybrid Tree-Based Twin-Bounded Support
Vector Machine-Based Model 305
15.3.3.1 Hybrid Tree Algorithm-C5.0 305
15.3.3.2 Twin-Bounded Support Vector
Machine-Based Model 307
15.4 Result and Discussion 309
15.5 Conclusion 313
References 313
Index 315

本书版权归John Wiley & Sons Inc.所有

 Foreword

In the dynamic landscape of today’s technological revolution, machine

learning and its applications span multiple domains, offering both oppor-
tunities and challenges. As we navigate this terrain, the significance of data
has shifted; it has transformed from merely passive entities to active driv-
ers influencing decisions, sculpting perceptions, and determining collec-
tive trajectories. This book serves a pivotal reference that sheds light upon
complex computational arenas and provides clarity to those navigating this
domain.
This book is more than an aggregation of knowledge. It epitomizes the
expertise and adaptability of current computational researchers and accen-
tuates the potential of metaheuristics. For those unfamiliar with the term,
envision metaheuristics as high-level strategists that steer a multitude of
heuristic methodologies toward their zenith. They offer the requisite tools
to address complex challenges where conventional algorithms might be
inadequate.
Throughout the book, you will find a wide range of applications and
potential uses of metaheuristics that span across domains from machine
learning to the cutting-edge fields of sustainability, communication, and
networking. It is fascinating to note that the algorithms aren’t just theoreti-
cal entities; they resonate with pressing real-world challenges. For instance,
consider the pivotal role of metaheuristics in life-saving applications like
breast cancer detection, or in ensuring security through anomaly identifi-
cation in surveillance systems and botnet attack detection.
Moreover, as we delve deeper, we witness the subtle yet profound syner-
gies between metaheuristics and contemporary technological innovations.
The chapters dedicated to the advancements in 5G and 6G communica-
tion, and the future of autonomous vehicles, are prime examples. These
sections underline the intricate balance and interdependence of the chal-
lenges we face today and the innovative solutions metaheuristics can offer.

xv

本书版权归John Wiley & Sons Inc.所有

 xvi Foreword

For researchers who dedicate their lives to exploration, practitioners

at the frontline of technological innovations, and students who look with
hopeful eyes toward the future, this book will be a pivotal tool. Let it guide
you, as it did for me, through the mesmerizing world of algorithms and
their real-world applications.

Diego Alberto Oliva

Universidad de Guadalajara, Mexico

本书版权归John Wiley & Sons Inc.所有

 Preface

While compiling this book, we were guided by a singular vision: to sculpt a

resource that seamlessly melds the theoretical intricacies of metaheuristics
with their myriad practical applications. Our aspiration was to produce a
reference that not only delves deeply into the subject, but is also accessi-
ble to readers across spectra, offering a holistic understanding that is both
profound and practical.
With every chapter, we strived to weave a narrative, oscillating between
the vast expanse of the topic and the intricate minutiae that define it.
The book commences with a foundational introduction, leading read-
ers through the labyrinthine world of metaheuristics. Going forward,
the narrative transitions, diving deeper into their multifaceted applica-
tions—spanning from the dynamic domain of machine learning to the
ever-evolving spheres of technology, sustainability, and the intricate web of
communication networks.
Metaheuristics present a promising solution to many formidable opti-
mization conundrums. Yet, their true allure comes not just from their the-
oretical promise but their practical prowess. This book attempts to unveil
this allure, transforming nebulous algorithms into tangible entities with
real-world resonances—whether in the life-saving realm of healthcare or
the cutting-edge world of vehicular communications.
We extend our endless gratitude to the brilliant authors, reviewers, and
countless others whose relentless dedication, insight, and expertise are evi-
dent in these pages. The editorial journey has been one of profound learn-
ing and growth for all involved. With each chapter, we have gleaned new
perspectives, and we hope this book becomes a wellspring of knowledge,
inspiration, and introspection for both scholars and professionals.

xvii

本书版权归John Wiley & Sons Inc.所有

 xviii Preface

In closing, we offer our sincere thanks to the Scrivener and Wiley pub-
lishing teams for their help with this book. We entreat you to immerse
your intellect and curiosity in the mesmerizing world of metaheuristics
and their applications. Here’s to an enlightening reading journey ahead!

Kanak Kalita
Narayanan Ganesh
S. Balamurugan

本书版权归John Wiley & Sons Inc.所有

 1
Metaheuristic Algorithms and Their
Applications in Different Fields:
A Comprehensive Review
Abrar Yaqoob1*, Navneet Kumar Verma2 and Rabia Musheer Aziz1
1
School of Advanced Science and Language, VIT Bhopal University, Kothrikalan,
Sehore, India
2
State Planning Institute (New Division), Planning Department Lucknow,
Utter Pradesh, India

Abstract
A potent method for resolving challenging optimization issues is provided by
metaheuristic algorithms, which are heuristic optimization approaches. They
provide an effective technique to explore huge solution spaces and identify close
to ideal or optimal solutions. They are iterative and often inspired by natural or
social processes. This study provides comprehensive information on metaheuris-
tic algorithms and the many areas in which they are used. Heuristic optimization
algorithms are well-known for their success in handling challenging optimization
issues. They are a potent tool for problem-solving. Twenty well-known meta-
heuristic algorithms, such as the tabu search, particle swarm optimization, ant col-
ony optimization, genetic algorithms, simulated annealing, and harmony search,
are included in the article. The article extensively explores the applications of these
algorithms in diverse domains such as engineering, finance, logistics, and com-
puter science. It underscores particular instances where metaheuristic algorithms
have found utility, such as optimizing structural design, controlling dynamic sys-
tems, enhancing manufacturing processes, managing supply chains, and address-
ing problems in artificial intelligence, data mining, and software engineering. The
paper provides a thorough insight into the versatile deployment of metaheuristic
algorithms across different sectors, highlighting their capacity to tackle complex
optimization problems across a wide range of real-world scenarios.

Keywords: Optimization, metaheuristics, machine learning, swarm intelligence

*Corresponding author: [email protected]

Kanak Kalita, Narayanan Ganesh and S. Balamurugan (eds.) Metaheuristics for Machine Learning:
Algorithms and Applications, (1–36) © 2024 Scrivener Publishing LLC

本书版权归John Wiley & Sons Inc.所有

 2 Metaheuristics for Machine Learning

1.1 Introduction
Metaheuristics represent a category of optimization methods widely
employed to tackle intricate challenges in diverse domains such as engi-
neering, economics, computer science, and operations research. These
adaptable techniques are designed to locate favorable solutions by explor-
ing an extensive array of possibilities and avoiding stagnation in subopti-
mal outcomes [1]. The roots and advancement of metaheuristics can be
traced back to the early 1950s when George Dantzig introduced the simplex
approach for linear programming [2]. This innovative technique marked
a pivotal point in optimization and paved the way for the emergence of
subsequent optimization algorithms. Nonetheless, the simplex method’s
applicability is confined to linear programming issues and does not extend
to nonlinear problems. In the latter part of the 1950s, John Holland devised
the genetic algorithm, drawing inspiration from concepts of natural selec-
tion and evolution [3]. The genetic algorithm assembles a set of potential
solutions and iteratively enhances this set through genetic operations like
mutation, crossover, and selection [4]. The genetic algorithm was a major
milestone in the development of metaheuristics and opened up new pos-
sibilities for resolving difficult optimization issues. During the 1980s and
1990s, the field of metaheuristics experienced significant expansion and
the emergence of numerous novel algorithms. These techniques, which
include simulated annealing (SA), tabu search (TS), ant colony optimiza-
tion (ACO), particle swarm optimization (PSO), and differential evolution
(DE), were created expressly to deal with a variety of optimization issues.
They drew inspiration from concepts like simulated annealing, tabu search,
swarm intelligence, and evolutionary algorithms [5].
The term “meta-” in metaheuristic algorithms indicates a higher level
of operation beyond simple heuristics, leading to enhanced performance.
These algorithms balance local search and global exploration by using ran-
domness to provide a range of solutions. Despite the fact that metaheuris-
tics are frequently employed, there is not a single definition of heuristics
and metaheuristics in academic literature, and some academics even use
the terms synonymously. However, it is currently fashionable to classify as
metaheuristics all algorithms of a stochastic nature that utilize randomness
and comprehensive exploration across the entire system. Metaheuristic
algorithms are ideally suited for global optimization and nonlinear mod-
eling because randomization is a useful method for switching from local
to global search. As a result, almost all metaheuristic algorithms can be
used to solve issues involving nonlinear optimization at the global level [6].

本书版权归John Wiley & Sons Inc.所有

 Metaheuristic Algorithms and Applications 3

In recent years, the study of metaheuristics has developed over time and
new algorithms are being developed that combine different concepts and
techniques from various fields such as machine learning, deep learning,
and data science. The development and evolution of metaheuristics have
made significant contributions to solving complex optimization problems
and have led to the development of powerful tools for decision-making in
various domains [7]. In order to find solutions in a huge search area, meta-
heuristic algorithms are founded on the idea of mimicking the behaviors of
natural or artificial systems. These algorithms are particularly valuable for
tackling problems that are challenging or impossible to solve using tradi-
tional optimization methods. Typically, metaheuristic algorithms involve
iterations and a series of steps that modify a potential solution until an
acceptable one is discovered. Unlike other optimization techniques that
may become stuck in local optimal solutions, metaheuristic algorithms are
designed to explore the entire search space. They also exhibit resilience
to noise or uncertainty in the optimization problem. The adaptability and
plasticity of metaheuristic algorithms are two of their main features. They
can be modified to take into account certain limitations or goals of the
current task and are applicable to a wide variety of optimization situations.
However, for complex problems with extensive search spaces, these algo-
rithms may converge slowly toward an optimal solution, and there is no
guarantee that they will find the global optimum. Metaheuristic algorithms
find extensive application in various fields including engineering, finance,
logistics, and computer science. They have been successfully employed in
solving diverse problems such as optimizing design, control, and manufac-
turing processes, portfolio selection, and risk management strategies [8].

1.2 Types of Metaheuristic Algorithms

We shall outline some of the most popular metaheuristic methods in this
section.

1.2.1 Genetic Algorithms

Genetic algorithms (GAs) fit to a cluster of metaheuristic optimization
techniques that draw inspiration from natural selection and genetics
[9–11]. In order to find the optimal solution for a particular issue, the core
idea underlying the GA is to mimic the evolutionary process. The genetic
algorithm has the capability to address challenges spanning various fields

本书版权归John Wiley & Sons Inc.所有

 4 Metaheuristics for Machine Learning

such as biology, engineering, and finance [12–14]. In the methodology of

the GA, a potential solution is denoted as a chromosome, or a collection of
genes. Each gene within the context of the problem signifies an individual
variable, and its value corresponds to the potential range of values that
the variable can take [15, 16]. Subsequently, these chromosomes undergo
genetic operations like mutation and crossover. This process can give rise
to a fresh population of potential solutions, resulting in a novel set of
potential outcomes [17–19].
The following are the major steps in the GA:

Initialization: The algorithm initializes a set of potential

responses first. A chromosome is used to symbolize each
solution, which is a string of genes randomly generated
based on the problem domain [20].
Evaluation: The suitability of each chromosome is assessed
based on the objective function of the problem. The quality
of the solution is evaluated by the fitness function, and the
objective is to optimize the fitness function by either maxi-
mizing or minimizing it, depending on the particular prob-
lem [21].
Selection: Chromosomes that possess higher fitness values
are chosen to form a fresh population of potential solutions.
Various techniques, such as roulette wheel selection, tourna-
ment selection, and rank-based selection, are employed for
the selection process [22].
Crossover: The selected chromosomes are combined
through crossover to generate new offspring chromosomes.
The crossover operation exchanges the genetic information
from the parent chromosomes and is utilized to generate
novel solutions [23].
Mutation: The offspring chromosomes are subjected to
mutation, which introduces random changes to the genetic
information. Mutation aids in preserving diversity within
the population and preventing the occurrence of local
optima [24].
Replacement: As the child chromosomes multiply, a new
population of potential solutions is formed and replaces the
less fit members of the prior population.
Termination: The technique proceeds to iterate through
the selection, crossover, mutation, and replacement phases

本书版权归John Wiley & Sons Inc.所有

 Metaheuristic Algorithms and Applications 5

Create initial Evaluate fitness of Apply

population each individual Selection

Crossover/Mutation

Termination Criteria
NO Reached?

Yes

END

Figure 1.1 Flowchart of the genetic algorithm.

until a specific termination condition is satisfied. Reaching

a predetermined maximum for iterations is one scenario for
termination, attaining a desired fitness value, or exceeding a
predetermined computational time limit.

The GA has several advantages, such as being capable of solving com-

plex issues, locating the global optimum, and being applicable to various
domains. However, the GA also has some limitations, such as the need for
a suitable fitness function, the possibility of premature convergence, and
the high computational cost for complex problems. Figure 1.1 shows the
flowchart of the genetic algorithm.

1.2.2 Simulated Annealing

Simulated annealing is a probabilistic method for optimizing complex
multidimensional problems by seeking the global best solution. It draws
inspiration from the metallurgical technique of annealing, which includes
heating and gradually cooling a metal to enhance its strength and resil-
ience [25]. Similarly, simulated annealing commences at an elevated tem-
perature, enabling the algorithm to extensively investigate a vast array of
possible solutions, and then slowly decreases the temperature to narrow
down the search to the most promising areas. SA works by maintaining

本书版权归John Wiley & Sons Inc.所有

 6 Metaheuristics for Machine Learning

a current solution and repeatedly making small changes to it in search of

a better solution. At each iteration, the algorithm calculates a cost func-
tion that measures how good the current solution is. The cost function
can be any function that assigns a score to a potential solution, such as
a distance metric or a likelihood function. Subsequently, the algorithm
determines whether to embrace or disregard a new solution by utilizing
a probability distribution that relies on the existing temperature and the
disparity between the costs of the current and new solutions [26]. High
temperatures increase SA’s propensity to embrace novel solutions even
if they have a higher cost than the current solution. This is because the
algorithm is still exploring the space of potential solutions and needs to
be open to new possibilities. As the temperature decreases, SA becomes
increasingly discriminating and admits novel solutions solely if they sur-
pass the existing solution. By employing this approach, SA prevents itself
from becoming trapped in local peaks and eventually achieves conver-
gence toward the global peak [27]. SA offers a notable benefit by effectively

Generate initial solution m and

calculate the objective function f(m)

Generate new solution m' and

calculate the new objective Δf=f(m')-f(m)
function f(m')

Yes No
Δf<=0

Accept the new solution m = Accept the new solution

m', f(m) = f(m') based on metropolis

No
Is the iteration
satisfied?

Yes
No
Is the stopping Decrease slightly the
criteria met? temperature

Yes

Terminate the computational process

and store the optimal solution

Figure 1.2 Flowchart of simulated annealing.

本书版权归John Wiley & Sons Inc.所有

 Metaheuristic Algorithms and Applications 7

addressing non-convex optimization problems, characterized by numer-

ous local optima. By permitting the acceptance of solutions with greater
costs, SA can navigate diverse areas within the solution space and prevent
entrapment in local optima. Moreover, SA boasts ease of implementation
and independence from cost function gradients, rendering it suitable for
scenarios where the cost function lacks differentiability. However, SA does
have some limitations. It can be slow to converge, especially for large or
complex problems, and may require many iterations to find the global
optimum. SA’s effectiveness is also influenced by the decision of cooling
schedule, which determines how quickly the temperature decreases. If the
cooling schedule is too slow, the algorithm may take too long to converge,
while if it is too fast, the algorithm may converge too quickly to a subopti-
mal solution [28, 29].
To put it briefly, simulated annealing is a highly effective optimiza-
tion method that has the capability to address intricate problems, multi-
dimensional problems with multiple local optima. It works by exploring
the solution space and gradually narrowing down the search for the most
promising regions. While it has some limitations, SA is a helpful tool for a
variety of optimization issues in the real world. Figure 1.2 shows the flow-
chart of simulated annealing.

1.2.3 Particle Swarm Optimization

Particle swarm optimization is a technique for optimization that employs
a population-based strategy to address a wide range of optimization prob-
lems. First introduced by Kennedy and Eberhart in 1995, this concept
takes inspiration from the coordinated movements observed in the flock-
ing of birds and the schooling of fish [30–32]. This algorithm emulates the
social dynamics exhibited by these creatures, where each member learns
from its own encounters and the experiences of its nearby peers, with the
aim of discovering the best possible solution [33]. The PSO method begins
by generating a population of particles, each of which acts as a potential
solution to the optimization issue at hand. These particles, which have
both a location and a velocity vector, are randomly distributed through-
out the search space [34]. The location vector represents the particle’s cur-
rent solution, whereas the velocity vector represents the particle’s moving
direction and magnitude inside the search space. Through iterative steps,
each particle’s location and velocity vectors undergo constant modifica-
tion and adjustment in the PSO algorithm, guided by its own best solution
encountered thus far and the solutions of its neighboring particles [35].
Collaborative learning continues until a predetermined stopping condition

本书版权归John Wiley & Sons Inc.所有

 8 Metaheuristics for Machine Learning

START

Initialization of t=t+1 t<T

control parameters Yes
No
Initialization of Update the velocity
particle position and & position of each The best position
velocity particle & the best
optimum is
obtained
Evaluate the fitness
Evaluation the fitness function of each
function for local and particle for local &
global best global best

END

Figure 1.3 Flowchart of the particle swarm optimization.

is met, such as when the desired outcome is attained or the maximum

number of iterations has been reached. Compared to other optimization
algorithms, the PSO algorithm boasts various advantages, including sim-
plicity, rapid convergence, and robustness [36, 37]. PSO has found applica-
tions in diverse problem domains, spanning function optimization, neural
network training, image processing, and feature selection. Nevertheless,
the algorithm does come with certain limitations. These include the risk
of premature convergence, where the algorithm may converge to subopti-
mal solutions prematurely, and challenges in effectively handling problems
with high-dimensional spaces [38].
In general, the particle swarm optimization algorithm is a robust and
effective optimization method capable of addressing numerous practi-
cal optimization problems. Its simplicity and intuitive approach make it
an appealing choice compared to more intricate optimization methods.
Figure 1.3 shows the flowchart of the particle swarm optimization.

1.2.4 Ant Colony Optimization

Ant colony optimization is a nature-inspired method that addresses diffi-
cult optimization problems by mimicking the behavior of ant colonies. The
program takes its cues from the behavior of ant colonies, specifically the
way ants communicate to discover the shortest path toward food sources.
The fundamental idea behind the ACO is to simulate the foraging behavior
of ants to solve optimization problems effectively. A simulated group of
ants is put on a graph representing the problem space in the ACO. These
ants navigate the graph by selecting the next node to visit based on the

本书版权归John Wiley & Sons Inc.所有

 Metaheuristic Algorithms and Applications 9

pheromone trails left behind by other ants. The strength of the phero-
mone trail represents the quality of the solution that passed through that
edge. As more ants traverse the same edge, the pheromone trail becomes
stronger. This is similar to how ants communicate with each other in real
life by leaving pheromone trails to signal the location of food sources [39,
40]. The ACO algorithm has several key parameters, such as the amount
of pheromone each ant leaves, the rate at which pheromones evaporate,
and the balance between exploiting the best solution and exploring new
solutions. The optimal values of the parameters in the algorithm are deter-
mined through a process of experimentation and refinement to obtain the
best possible results for a specific problem [41].
The ACO has showcased impressive achievements in resolving diverse
optimization challenges, including but not limited to the traveling sales-
man problem, vehicle routing, and job scheduling. One notable advantage
of the algorithm is its ability to swiftly discover favorable solutions, even

Randomly Ant
Initial Population
Located

Each Ant
Randomly
Choose a path

Update
Pheremone Trial

Search Process
Pheremone Continue
Evaporation

Evaluation Stop
Condition

Results

Figure 1.4 Flowchart of the ant colony optimization.

本书版权归John Wiley & Sons Inc.所有

 10 Metaheuristics for Machine Learning

when confronted with extensive search spaces. Furthermore, because the

ACO belongs to the category of metaheuristic algorithms, it can be applied
to a variety of situations without requiring a deep understanding of the
underlying structure of those problems [42]. Figure 1.4 shows the flow-
chart of the ant colony optimization.

1.2.5 Tabu Search

The tabu search is a metaheuristic technique utilized for optimization prob-
lems, initially proposed by Fred Glover in 1986. It has gained significant pop-
ularity across diverse domains, including operations research, engineering,
and computer science. The core concept behind the tabu search involves
systematically traversing the search space by transitioning between different
solutions in order to identify the optimal solution. However, unlike other
local search algorithms, the tabu search incorporates a memory structure
that records previous moves executed during the search. These data are then
used to steer the search to potential places within the search space [43]. The
tabu list, a memory structure that plays an important part in the algorithm, is
at the heart of the tabu search. This list serves to store and remember previous
moves made during the search process, ensuring that the algorithm avoids
revisiting solutions that have already been explored. By utilizing the tabu list,
the tabu search effectively restricts the search to new and unexplored regions
of the solution space, promoting efficient exploration and preventing repet-
itive or redundant searches. This list is used to enforce a set of constraints,
known as the tabu tenure, which determines how long a move is considered
tabu. By imposing this constraint, the algorithm is compelled to investigate
diverse regions within the search space and evade being trapped in local
optima. This ensures that the algorithm remains dynamic and continually
explores new possibilities, preventing it from being overly fixated on sub-
optimal solutions [43]. The tabu search is a versatile optimization algorithm
applicable to both continuous and discrete optimization problems. When
addressing continuous optimization problems, the algorithm typically uses a
neighborhood function to generate new solutions by perturbing the present
solution. In the event of discrete optimization problems, the neighborhood
function is typically defined in terms of specific moves that can be made to
the solution, such as swapping two elements in a permutation. The effective-
ness of the tabu search is based on a number of variables, such as the choice
of neighborhood function, the tabu tenure, and the stopping criterion. The
algorithm can be enhanced by using various strategies, such as diversifica-
tion and intensification, which balance the search space’s exploitation and
exploration [44].

本书版权归John Wiley & Sons Inc.所有

 Metaheuristic Algorithms and Applications 11

The tabu search, a metaheuristic approach introduced by Fred Glover

in 1986, is utilized for optimizing problems. Its application spans a wide
array of domains, including operations research, engineering, and com-
puter science, establishing it as a widely recognized technique. The funda-
mental principle of the tabu search involves a systematic traversal of the
solution space, shifting between different solutions to ascertain the optimal
one. Distinguishing itself from conventional local search algorithms, the
tabu search incorporates a memory structure that logs prior moves exe-
cuted during the search process. This stored information guides the search
toward unexplored areas within the solution space [45]. Central to the tabu
search is the tabu list, a pivotal memory structure. This list retains and
recalls previous moves executed during the search, ensuring that revisit-
ing already explored solutions is avoided. The tabu list effectively restricts
the exploration to untrodden regions, preventing redundant searches and
fostering efficient investigation. Governing the tabu list is the concept of
the tabu tenure, setting the duration for which a move remains prohib-
ited. This constraint compels the algorithm to explore diverse solution
space regions, eluding entrapment in local optima. This dynamic approach
continuously explores novel avenues, counteracting fixation on subopti-
mal solutions [46]. The tabu search is a versatile optimization algorithm
suitable for both continuous and discrete optimization challenges. For
continuous optimization, a neighborhood function is commonly used to
generate new solutions by perturbing the current one. In the context of
discrete optimization, the neighborhood function is typically defined by
specific permissible moves, such as element swaps in a permutation. The
efficacy of the tabu search hinges on factors like neighborhood function
choice, tabu tenure, and the stopping criteria. The algorithm can be aug-
mented through strategies like diversification and intensification, ensuring
a balance between exploiting and exploring the search space.
In general, the tabu search is a robust and adaptable optimization
technique that has demonstrated its effectiveness in addressing diverse
problem sets. It can be employed independently or integrated into more
intricate optimization algorithms. Its popularity stems from its versatility
and straightforwardness, making it a favored option for tackling real-life
challenges in various domains.

1.2.6 Differential Evolution

The DE is an optimization algorithm based on populations, originally cre-
ated by Storn and Price in 1997 [47]. It fit to the category of evolution-
ary algorithms that iteratively grow a population of potential solutions to

本书版权归John Wiley & Sons Inc.所有

 12 Metaheuristics for Machine Learning

find the optimal solution. The algorithm adheres to the fundamental steps
of mutation, crossover, and selection, which are key elements commonly
shared among numerous evolutionary algorithms [48].
In the process of the differential evolution, a population of potential
solutions undergoes iterative evolution through the implementation of the
following sequential steps:

1. Initialization: A population of N possible solutions is pro-

duced at random.
2. Mutation: Involves randomly selecting three candidate solu-
tions and modifying them to create a trial vector.
3. Crossover: It is a technique used in optimization algorithms
to create a new candidate solution by combining the trial
vector with the target vector.
4. Selection: If the new candidate solution has a higher fitness,
it will take the place of the target vector.

The success of the differential evolution depends on the selection of the

optimization technique’s adjustable settings, such as mutation rate, cross-
over rate, and population size [49, 50]. Several variants of the DE have
been proposed, including the SHADE (success history-based adaptive dif-
ferential evolution) and JADE (adaptive differential evolution) algorithms,
which incorporate adaptive control parameters to improve the algorithm’s
performance.

1.2.7 Harmony Search

The harmony search (HS) is an optimization technique motivated by the
musical improvization process. Geem [125] had put forward the idea in
2001 and has now been used to solve several optimization issues. The tech-
niques mimic the process of improvization by a group of musicians, where
they adjust their pitches (or notes) to create harmony. In the HS, the choice
variables of an optimization problem in high school are comparable to
musical notes, and the value of the goal function reflects harmony [51].
Starting individuals of decision variable vectors (i.e., the notes) are used
in the HS approaches and iteratively search for better solutions by generat-
ing new solutions through the following steps:

1. Harmony memory: A set of the best candidate solutions (i.e.,

the harmonies) is maintained.

本书版权归John Wiley & Sons Inc.所有

 Metaheuristic Algorithms and Applications 13

2. Harmony creation: A potential solution is created by chance

selecting values from the harmony memory.
3. Pitch adjustment: The values in the new candidate solution
are adjusted with a probability based on a pitch adjustment
rate.
4. Acceptance: The new candidate solution is accepted if it
improves the objective function value.

The control variables, such as the harmony memory size, pitch adjust-
ment rate, and number of iterations, affect how well the HS performs. The
method follows the core phases of mutation, crossover, and selection, and
the approach has been utilized to tackle diverse optimization challenges,
such as managing water resources, designing structures, and operating
power systems [52, 53].

1.2.8 Artificial Bee Colony

The artificial bee colony (ABC) is a population-based optimization method
that draws inspiration from honey bees’ feeding habits. Since its introduc-
tion by Karaboga in 2005, the technique has been used to solve a number
of optimization issues [54]. The ABC mimics the foraging process of bees,
where they search for food sources by visiting the flowers in the vicinity of
the hive [55].
The artificial bee colony technique starts with an arbitrarily gener-
ated population of candidate solutions (i.e., food sources) and iteratively
searches for better solutions by simulating the foraging process of bees
through the following steps:

1. Phase of employed bees: The employed bees develop new

candidate solutions by modifying the values of current
solutions.
2. Phase of the onlooker bees: The onlooker bees choose the
candidate solutions with the highest fitness values and send
this information to the employed bees.
3. Phase of scout bees: The scout bees search for new candidate
solutions by randomly generating new solutions.

The success of the ABC depends on the control parameters, such as

the population size, the number of iterations, and the probability of aban-
doning a food source. The algorithm has been applied to a wide range of

本书版权归John Wiley & Sons Inc.所有

 14 Metaheuristics for Machine Learning

optimization problems, including image processing, wireless sensor net-

works, and fuzzy control systems [56].

1.2.9 Firefly Algorithm

A metaheuristic optimization technique called the firefly algorithm (FA) is
based on how fireflies behave. Dr. Xin-She Yang first presented the FA in
2008. The social behavior of fireflies, which is characterized by their flash-
ing light to attract mates or prey, served as an inspiration for the algorithm
[57]. The firefly technique’s primary role is to increase light intensity, and
in order to maximize light intensity, fireflies travel toward the brighter fire-
flies. Starting with a random population of fireflies, using the goal function,
the program calculates the light intensity of each firefly [58]. The move-
ment of fireflies is governed by their attractiveness, which is determined
by the brightness of their light, and the distance between them. The fire-
flies move toward the brighter fireflies and update their positions until the
maximum light intensity is achieved [59]. Numerous optimization issues,
including those involving machine learning, image processing, and func-
tion optimization, have been effectively solved using the firefly algorithm.

1.2.10 Gray Wolf Optimizer

The gray wolf optimizer (GWO) is a metaheuristic optimization system
based on wolves’ social structure and hunting techniques. The algorithm
was developed in 2014 by Seyed Ali Mirjalili, Seyed Mohammad Mirjalili,
and Andrew Lewis [60]. The optimization issue is viewed in the gray wolf
optimizer as a wolf pack’s prey–predator dynamic. The wolf population
used by the method is randomly divided into four groups: alpha, beta,
delta, and omega. While the omega wolf is the weakest and has the low-
est fitness value, the alpha wolf assumes the role of the leader within the
wolf pack and has the greatest fitness value [61–63]. The movement of the
wolves is governed by three different types of hunting techniques: hunting
for prey, following the alpha wolf, and surrounding the prey. The wolves
adjust their locations based on these techniques until the optimal solution
is accomplished. The gray wolf optimizer’s application has been expanded
to address a variety of optimization issues, such as function optimization,
engineering design, and feature selection [64, 65].

1.2.11 Imperialist Competitive Algorithm

The imperialist competitive algorithm (ICA), a metaheuristic optimization
method designed to address numerous optimization issues, was inspired

本书版权归John Wiley & Sons Inc.所有

 Metaheuristic Algorithms and Applications 15

by the political and economic rivalry of imperialist nations. Amir Hossein

Gandomi and Ali Alavi introduced the method in 2010 [66]. In the impe-
rialist competitive algorithm, the optimization problem is considered
as a competition between empires. The algorithm starts with a random
population of empires, which are categorized into two groups: imperialist
and colonies. The imperialist empires have higher fitness values, and they
expand their territories by annexing colonies. The movement of empires is
governed by two different types of actions: assimilation and revolution. In
the assimilation process, the imperialist empires try to improve the fitness
of their colonies, while in the revolution process, the colonies rebel against
their imperialist empires and become independent. Various optimiza-
tion problems, including function optimization, image segmentation, and
parameter estimation, have been effectively solved through the successful
application of the imperialist competitive algorithm [67, 68].

1.2.12 Bat Algorithm

Xin-She Yang created the bat algorithm (BA) in 2010, which is a meta-
heuristic optimization technique used to tackle a variety of optimization
challenges [69]. The bat algorithm is inspired by bats’ echolocation activity,
which uses ultrasonic noises to navigate and locate prey in the dark. It rep-
licates the activity of bats in their hunt for prey in order to discover the best
solution to a given optimization issue [70].
To solve an optimization problem using the bat algorithm, a population
of bats is created in the search space with random placements and veloc-
ities. The bats move randomly, emitting frequencies proportional to their
fitness values. Bats with better fitness emit higher-frequency sounds that
attract other bats toward their position in the search space [71]. In adding
to the frequency-based attraction mechanism, the bat algorithm includes a
random walk component that allows the bats to explore uncharted regions
of the search space. During each iteration, the algorithm updates the veloc-
ity and position of each bat using information on the best solution found
thus far, as well as the loudness and frequency of its emitted signal. The
algorithm iterates until it reaches a present stopping point, such as a maxi-
mum number of iterations or a goal fitness value [72].

1.2.13 Cuckoo Search

The cuckoo search, a metaheuristic optimization method inspired by the
cuckoo bird reproductive behavior, was introduced in 2009 by Xin-She
Yang and Suash Deb. Cuckoo birds use an unusual approach in which they

本书版权归John Wiley & Sons Inc.所有

 Another Random Document on
Scribd Without Any Related Topics
There was a noticeable pause; then Guninana answered, trying to
make her voice sound ordinary. “She went to play and has not come
back yet.”
“But the others are back long ago!”
“She went into the woods.”
Taptuna’s voice sounded rough for his proud soul was full of alarm
which he would have liked to keep from Okak.
“Ah—into the woods—and she has not returned.”
Each slow word was a knife twisted in their hearts. Dead silence
followed. It is not necessary to talk when all know what the others
are thinking. At last Okak broke out violently:
“This is exactly what I expected! We had better rouse the village,
neighbor, and go in pursuit.”
His use of that final strange word stabbed his belief home.
“Nonsense!” protested Taptuna, but the familiar exclamation lacked
force. It seemed to drop away into darkness. Okak’s voice continued
harshly:
“Ah, yes! You have been saying ‘Nonsense, nonsense’ to me all
summer. But now this is not such ‘nonsense’ if the Indians have
taken Noashak. And why should we suppose they haven’t got her?
Has any child ever strayed from the camp before? Not one! Certainly
they have enough intelligence to return if they are not prevented.
And what else could prevent her—who else but your precious red
traders! It is fortunate if they have only carried her away, and have
not already taken her teeth for their children’s toys and her hair as a
decoration.”
“Don’t!” Guninana cried shuddering.
Though his speech was cruel she knew Okak as a faithful friend. He
had already put on his stoutest pair of boots and was selecting his
best arrows with trembling hands.
“Where is Omialik?” he asked.
“Hunting.”
“It is as well for him that he is hunting!”
This threat sounded so sinister the others were quite taken aback.
They had not expected blood and vengeance of the timid Okak.
Seeing Taptuna hesitated the little man took another tone, urging:
“Come, neighbor, there is no time to lose. A volunteer party must
start for the Indian encampment at once.”
When one person makes up his mind about anything so very
positively, he is apt to carry conviction to others. Taptuna did not
know what to think. Okak’s turning into a man of action was an
uncanny business in itself. It made him feel as you would feel if a
statue on the street corner suddenly came to life and commenced
issuing orders. Circumstances seemed to prove his fears and hatred
just. They had held the thought of Indians from the first, though
unconfessed; and nothing came to mind to overthrow their
neighbor’s reasoning. Besides, both realized that neither Okak nor
the village knew the worst—the fact of Jimmie Muskrat’s trickery.
“Perhaps—perhaps! It will be better to go down and see—and be
sure,” Taptuna muttered.
“Anything is better than nothing. Do something!” the mother
moaned.
At that her tall, competent husband turned and meekly followed his
fussy companion across the open ground to the mottled tents
looking so much like rocks under the pale radiance of the autumn
moon.
Kak awakened to the menace of an empty village, deserted work, his
mother’s grief, and the frightened faces of the women who had
come to sympathize. Okak’s accusations had convinced them. They
told the boy without a shadow of doubt that Noashak had been
carried off by Indians and the men were gone after her. All this
tragedy springing out of his one moment’s ill nature was more than
Kak could stand. It seemed very unfair. Nobody spoke a word of
blame, but he felt they all knew it was his fault, and unable to meet
their looks he stole away and hid amid the underbrush till the search
party should return.
When he heard them coming he crept out hopefully. But the worst
news was already leaping from lip to lip by the time he got home.
They had found the camp site but no campers. The remains of the
lodges were freshly deserted, and it was all too evident the Indians
had run away with their prize. Taptuna, nearly crazy, had insisted,
against his people’s advice, on immediate pursuit. He would have
started alone had not the Kabluna’s two Eskimos volunteered to go.
The three were following hot on the redskins’ trail.
Kak revisited the underbrush and gave himself up to despair. He had
felt remorseful last night; now his heart sank into his very boots.
Omialik being away added the last drop of bitterness to the cup. This
distress was purely unselfish. Much as the boy longed for advice and
comfort, he really wanted his friend to come back and clear his own
good name. Women in the village were already telling how the white
man had been party to the whole plot; asking, “Aren’t his Eskimos
glad for an excuse to escape?” They said Omialik would never come
again, would never dare to show his face. This hurt Kak as nothing
else could have done. It was difficult to keep doubt out of his valiant
little soul when doubt seethed all around him. Of course he did not
believe their lies, but the sting and strain of loyalty which stands
against the mob, and the soreness which endurance leaves in the
human heart are fierce emotions for a child. Kak writhed in double
torture; then gradually his mood shifted from crushed humiliation to
stern resolve. Since it was his fault Noashak had fallen into the
Indians’ hands, it was plainly his duty to rescue her; and it was his
privilege to defend Omialik—to warn him.
Lying on his back, staring up into the blue sky, Kak thought it all out
carefully. He would go after his sister. No need to waste time
scouting around by the deserted camp, he could strike boldly across
country till he reached the eastern end of Great Bear Lake, and once
there he would find Mr. Selby. If Mr. Selby proved friendly and asked
the Indians living about him to help, then Kak would be able to send
a warning to Omialik, for his friend must know his plans.
Fired with ambition the boy crept back to their tent, made up a small
package of dried meat, took his bow and arrows—all his new ones
that he had been so eagerly laying aside for use on the homeward
journey—and stole away.
Guninana sat among the neighbors in the largest tent, where a
shaman in a sort of trance, with wild contortions and weird words,
sought Noashak. Kak kept out of it. He did not want to be stopped
and questioned. “Mother will understand when she sees I have
taken these arrows,” he thought, as he ran on silent feet down the
nearest path. Kak too looked like a deer in his deerskin clothing. The
trees held out welcoming arms, and the rocks were mottled with
grays and browns. In a few minutes the wilderness swallowed him,
leaving no trace.
He struck boldly south. The forest consisted mostly of slender spruce
in scattered formation, so at first he made good progress. But when
he had gone perhaps six hours’ travel the woods grew denser; thick
enough to try both his strength and patience. He was thinking about
making camp, sitting down for a rest and a bite of food anyway,
when a rustle in the branches set his pulses throbbing. The forest
lay still but not silent; a light wind from the north, sighing
continuously, swayed the tapering tree tops; but this noise he heard
was different from any wind noises—a persistent rustling through
the alders. It was sunset and darkening here in the woods, and poor
Kak, who had been like a lion a moment before, felt all the courage
oozing out of him. He fell on one knee behind a log. The sound
came nearer, grew unmistakable. Some large body moved through
the copse. The young hunter laid an arrow across his bow and
waited with every muscle taut. On it came, near—so near he began
to tremble for his safety. What if a grizzly bear loomed suddenly out
of the dusk above him! The boy knelt trembling, with distended eyes
riveted on that spot where the stealthy noise seemed to approach.
“Whatever it is, it’s coming so close I can’t miss,” he thought, and
bent the bow. Swiftly the bushes parted, letting a dark mass tower
over him. It stood with its back to the waning light and might easily
have been an animal by its shaggy outline; but Kak saw. His muscles
relaxed in sickening reaction as the human form sprang at him over
the log and seized his arm.
“Good gracious! Don’t kill me!” cried a familiar voice.
“Omialik!”
Two sorts of relief rang in that cry. The Kabluna was on his way back
—then they had all told lies, lies, lies! The boy’s sorrowing heart
rushed out to his friend, whom he had so nearly shot; he threw
himself into the white man’s arms and cried like a baby.
“Why, Kak! Why, Kak! Were you lost? Were you scared?”
Omialik repeated over and over as he patted the sobbing youngster:
“Brace up. It’s all right now. We’re not many hours from home.
Come—come! Brace up.”
“It isn’t me,” cried the boy. “It’s Noashak. She’s been stolen by
Indians!”
“What nonsense!”
“That’s what dad said, but she’s gone just the same. The men went
down to the Indians’ camp to hunt for her; but the Indians are gone.
And you were gone too! The women are telling that you were in
league with Muskrat.”
“Great Jehoshaphat!”
This was startling news—bad news—bad enough to make the white
man want to hear it quite correctly.
“They’ve been to the camp, you say, and found the Indians gone?”
“Yes, and father is following with your Eskimos: the rest of the
search party came home.... It is all my fault Noashak’s lost. She ran
away into the woods because I was cross with her; so I thought I’d
better try and bring her back. And I was going to the lake to leave a
message with your Selby about how mad the village is, so—so that
you wouldn’t go there without your gun.”
“You intended to warn me? That was kind.”
Omialik’s eyes grew soft. One glance at his face was sufficient
reward for Kak. Look and words together acted like balm on the
boy’s bruised self-esteem. As he sat by his friend, eating dried meat
and telling him every detail of their scare, his spirits rose. It seemed
possible Noashak had never been near those deserted lodges—that
they might all have been wrong. And he was prepared to accept the
white man’s judgment when it came.
“I don’t believe Muskrat had anything to do with this business. It
would be best, my lad, for you and me to return to the village and
set matters right there. If your father is not back—if they have no
news—we can start systematic search instead of running off on a
wild goose chase. Maybe the child is only lost. What made you so
sure she was stolen?”
Kak thought hard. “The women told me so,” he answered. “And
Okak told them so. He was positive.”
Omialik smiled. “Okak was always crazy-frightened of Indians.”
“But what he said is true. Noashak would certainly come home from
her play unless something was keeping her. The kids never go far.”
“Well, something else might have prevented her. Suppose she had
fallen, or——”
“Don’t!” cried her brother in the same tone Guninana had used. “I’d
rather it was Indians than animals!”
The boy found himself suddenly, vividly, plunged back into that
terrifying moment before Omialik appeared, when his courage oozed
out of him, his hair stirred on his head, and cold sweat started from
every pore. He tried to imagine his little sister so amazed,
surrounded, trapped by some wild beast of the woods—but it was
too awful.
“Come on!” he cried, springing to his feet. “We’ve got to get ahead!”
They had been talking a long time and it was now dark with a cloudy
sky. The white man’s instinct was to camp and wait for daylight. But
Kak urged him so to “Come along,” to “Try,” that he gave in against
his better judgment, and they began scrambling through the thick
brush. It was slow, heavy travel and after an hour’s effort, Omialik
stopped.
“No use, Kak, we are only losing our way and getting all mixed up. I
haven’t any idea which way we are heading. This seems a likely
spot, so far as one can feel, and I hear water. Let us camp and wait
for morning.”
Kak was about ready to drop from fatigue and silently agreed. They
built a little fire for the night was cold, and ate some more dried
meat, drinking great refreshing draughts from the spring which
Omialik’s quick ear had not failed to note.
“What is that strange smell?” asked the boy, sniffing the keen,
autumn wind.
“Caribou, or I’m mistaken. My, but it’s strong! We must be close to
an enormous herd—the first caribou I have struck in three days, and
it’s so pitch-black I can’t see my hand before my face! What rotten
luck!”
“Well, I’m glad it is dark! I’m too tired for hunting,” Kak answered,
and throwing himself on a bed of moss, immediately slept.
The young hunter awakened in the early morning of a quiet lowering
day. Caribou scent hung heavy in the still air. He noticed a strange
vibration through the ground, heard the thud and rustle of trotting
feet; sat up and shook his companion.
Omialik rolled over sleepily, opened one eye, grew conscious also of
that odd trembling in the ground, opened the other eye, and lay
staring into the clouds.
“Whatever is it? Do you feel—do you hear?” asked Kak in excited
whispers. “Yes, and I smell it too!”
The Kabluna rose on one elbow. “Must be caribou traveling,” he said.
“A large band—an immense band!... Listen to the ripple of their
feet.... Wonderful! Let’s get out of here to some place where we can
see.”

THE WHOLE PLACE SEEMED TO BE A MOVING RIVER OF

DEER.

He scrambled up and pushed through the copse, Kak following. It

might have been an eighth of a mile to where the trees thinned.
There an unbelievable sight met their eyes. Caribou were marching
past in solid columns, two, three, or more abreast. These columns
were only a few yards apart and extended as far as eye could see
into the sparse woods. The whole place seemed to be a moving river
of deer.
“I wonder how long it’s been going on,” the white man exclaimed.
“My word, I’m glad we wakened before they all passed! I wouldn’t
have missed this sight for anything!”
They stood there a long time waiting, expecting the herd to peter
out, its spectacle coming to a sudden stop like a battalion marching
by. But the solid columns continued to pour on—the river flowed and
flowed.
“Marvelous!” sighed Omialik.
“Perhaps we can get along through the woods,” Kak suggested; for
the fascination of the marching host paled a little when he
recollected his sister. The white man could not bear to tear himself
away. This was the grandest exhibition of the riches of the north he
had ever seen. He wanted to look and look, convincing himself of its
reality, so that when he returned to his own country and people
talked about “those cold waste regions,” and “the barren Arctic,” he
could remember this and say: “You are all wrong. Hundreds of
thousands of animals roam over that so-called desert; birds and
butterflies and insects, millions of insects, infest it; and caribou
travel there by regiments.” Noashak’s peril left him no choice but to
turn his back on the deer. They tramped through the copse where
they had slept. In its thickest part the sound of running feet died
down a little, then it swelled again, grew sharper, more distinct.
“Foxes!” cried Kak. “I believe we’re coming on another lot over
here!”
It was so, their copse proved to be an arm of the forest thrusting
itself thickly down along either side of a small stream. And they
broke out of it suddenly, opposite their first stand, to find more solid
columns of migrating deer moving steadily past. These animals
walked as close one after the other as possible, while row beyond
row lined all the visible area.
“Aren’t you hungry?” Omialik said. “Shall I kill some fresh meat for
breakfast?”
“First rate!” Kak answered. Then glancing at the closely packed
animals, “But it seems a kind of shame!”
“Good for you! That is the right sporting spirit, my boy; stalk your
game, don’t have it driven. However, necessity is master here—and I
don’t believe one will be missed. What a chance this to kill a whole
winter’s food supply! If only my men and your dad were along to
help us build caches. It would be waste to slaughter the poor things
and leave them for wolves.” Omialik stood watching, then he glanced
at his companion. “Suppose you do the shooting this time and save
ammunition.”
Excitement fluttered up the boy’s nerves; he only hoped he did not
show it as he anxiously selected one of the new arrows and bent his
bow. Kak had never killed a deer, and there was little glory, he knew,
in killing at such easy range; yet he got a thrill when the large buck
he had picked staggered and fell among the herd. Omialik’s praise
was sweet.
They built a fire and feasted on roast ribs, making a quick meal of it,
for Noashak’s little figure seemed always to be flitting before Kak’s
eyes.
As the caribou were now moving against a shifted wind, almost
directly away from the village, the man and boy were able to walk
between two columns when chance offered breaking through one
line into the space which divided it from the next, walking there
awhile, and at the first opportunity repeating the maneuver; always
keeping to the right and slowly working out of the herd. After they
had left behind the last straggling groups, a couple of hours’ fast
travel brought them home.
By late afternoon, as they neared the village, the brother began to
worry. “We won’t have much daylight for searching,” he grumbled,
“and I know how it will be, everybody crowding around gabbing,
trying to get in a word with you or at you—delaying us no end.”
The white man was endeavoring to cheer him by promises of a
speedy departure; when who should come running to meet them but
Noashak herself.
Kak’s throat choked up at seeing her. “What happened?” was all he
could say.
The little girl seized Omialik’s hand and jumped around and rubbed
on him in quite her old, bothersome manner.
“Don’t act so much like a chipmunk. Come. Tell us your story!” He
laughed at her mauling, and captured both small hands in one large
glove. “What happened after you ran away to play with the hares
and marmots?”
“I wanted to go right off where Kak would have a lot of trouble
finding me, because he was mean. You were mean, Kak! I ran and
ran till I was so tired I lay down—maybe I had a little nap. When I
felt rested and thought you had been looking for me long enough I
tried to go home; but the sun hid behind clouds and I didn’t know
which way was home, and still I kept on going. Then numbers of
caribou came feeding near by—more and more and more. It began
to grow dark and I cried. That didn’t stop the darkness a bit; so by
and by I ceased crying and looked around for a bed. There was a
nice, low island of rock with three spruce trees growing on it, and
smooth ground all covered with moss, and I thought: ‘That will make
me a fine house.’ With such a lot of animals around I wanted a safe
place. I climbed up. It was almost dark and the night grew blacker
and blacker for a while; but presently the clouds blew away, and the
stars shone and the moon. There was an awful smell and the sound
of many animals running. I could see antlers like trees rushing past,
and the wolves howled, and——”
“You were scared and howled with them.”
“Yes, I did,” the child answered boldly. “I cried myself to sleep. When
I woke up it was bright day and the whole world was covered with
caribou—such lots and lots of caribou, all going in the same
direction! There were wolves among them and I was frightened to
go into the herd, so I sat still and waited. I was on the island with an
ocean of deer rushing by. They kept me on the island. I had nothing
to eat but berries, and I cried and hoped you would soon come to
find me.”
It was so. That day the child had lain alone on the dry, vibrating
ground under low clouds, and watched the cold, blue evening fall;
while those gray, shadowy, moving legs and tossing, antlered heads
came on, and on, and on. The thud, thud of running roofs made a
strange lullaby. The wind had risen to a sighing moan, and now that
night drew in wolves, racing with the herd, howled dismally.
All through the darkness deer continued trotting by, and to the
tramp and tremble of their small, innumerable feet Noashak waked a
second time.
She felt very lonely and sad as well as hungry, and scarcely thought
it worth while to sit up and look at those interminable creatures.
Imagine her joy, then, on finding one edge of her rock quite free—
luckily for her the edge toward home. This was because the breeze
had shifted, making the caribou, which usually travel into the wind,
alter their course. Gradually, while the captive slept, the columns
had bent westward till the whole, vast herd was swinging down on
the far side of her island. The instant she took it in Noashak jumped
up and hurried out of prison.
“I’ll never, never again be so naughty as to run away!” the child
promised, shaking her head violently; but her seriousness lasted
only five seconds.
“What do you think?” she cried, hopping on one foot. “Okak said
Indians had carried me off. I wish they had! Then I could have seen
their lodges, and I wouldn’t be back till father saved me, and killed
Jimmie Muskrat; and everybody would still be scared.”
“What! Do you like to frighten us, you mischief?”
“Course I do! It’s lots of fun. Being away is tiresome; but it’s grand
getting home! Everybody gives you things—see. Here’s Okak’s charm
against evil.” She held up a dried bumblebee hung in a bag on a
sinew about her neck. “Mother says I look too much like a fawn and
she has promised to make me a coat with bright red trimmings if we
can get the ocher at Cape Bexley. Do you hear, Kak? I’m to have a
new red coat! It’s so I shall never get lost any more. But I’d like to
be lost sometimes if I could see all those caribou. Nobody believes I
did see them. They say I dreamed it—but I really and truly did.”
“Bully for you! Stick to it,” Kak cried. “They were real, all right, and
you saw them. Don’t let the villagers humbug you out of that. We
saw them, too, and we killed one and ate it—that’s proof it was
real!”
“Only one, worse luck!” Omialik exclaimed. “But now you are safe,
miss, we’ll hurry back and lay in some meat. Where is your father?”
he asked; for there would be need of all hands to skin and cut up
the deer.
“Dad’s still looking for me, and your Eskimos are with him. I guess
they’ll be pretty anxious by now.... Oh, I do hope they’ll come here
soon so we can start to Cape Bexley—I do want my little red coat!”
 CHAPTER X
Homeward Bound
All very well to talk so lightly about going to Cape Bexley; when it
really came to the point, leaving meant taking leave and this was a
bad business. Kak’s heart broke, for his friend, Omialik, stayed
behind. It was the white man’s intention to return down Horton
River to Franklin Bay and go from there to Banks Island—a long and
dangerous journey into the unknown. The boy burned to accompany
him.
“Later on, later on, when your legs are a bit longer for walking,” the
explorer promised.
Kak tried to smile, tried not to show the hollow feeling this
separation planted in the pit of his stomach; but it took moral force.
He gulped.
“Brace up, old chap.” The Kabluna patted his shoulder. “I’m coming
back, you know. You will see me in Victorialand again—unless by
then you have gone to Herschel Island to learn to shoot.”
Talking about impossible dreams as if they were bound to happen
makes them seem jolly real. Kak managed to choke back his sorrow,
and freshly convinced that life was a grand adventure, ran after the
party who were already trekking north.
Crossing the prairies with all their gear and trade goods, the wooden
dishes, pails, lamp supports they had made, and pieces of rough
wood piled on the sled, proved an entirely different experience from
their tiresome, hot, hungry tramp southward. The new sled ran
lightly on snow ample to cover the ground and not too heavy for
walking. Taptuna was careful to pack a good supply of food, and
halfway across the tundra they found their old cache. All laughed
heartily to think how much worse they had needed it in the summer
than they did now.
With favorable weather and little time lost hunting they made a
record trip. Spirits mounted at every mile. Guninana sniffed the
ocean air joyfully and said how fine it would be to live in a
comfortable snow house, away from buzzing flies and boiling hot
sun, and that perpetual sense of work always awaiting them in the
woods.
Frost made Kak feel like a war horse. He longed to have the flat ice
under his feet again, with two dogs, perhaps three if he was lucky,
harnessed to the sleigh, and run—run—run—abandoning himself to
that glorious sense of space and motion which was his heritage.
The first person he hunted up at the cape was Kommana.
“Got that pup for me?” he shouted.
“Got that snow shovel?”
“Sure thing!”
Kak proudly produced their wonderful slab of spruce, and when
everybody about had admired and praised it he was offered his
choice of the six-months-old dogs.
The boys’ fathers were party to this trade, for a single piece of wood
the width of the one they had brought was considered very valuable
—worth almost as much as Taptuna’s new sled.
This was a large village, many Eskimos from the north and east had
come to trade, and things took on the character if not the
appearance of one of our small-town fairs. Besides their business the
traders indulged in sports, jumping and racing and playing football.
Their balls are made of soft leather sewed together in sections,
much like ours, and are stuffed with caribou hair. The hair of the
caribou, being hollow, is very buoyant; this is why the animals float
nearly half out of water after they are killed. Their hollow hair is
often used in manufacturing life preservers and is considered better
than cork. Balls filled solidly with it bounce quite well, and the
Eskimos have a lot of fun kicking them about. Kak was rather good
at games, though, of course, he could not hold his own against men,
but Kommana had no use for them.
“You’ll be old before you’re grown!” Taptuna jollied him. “Come and
take a turn at this—just try.”
He sent the ball spinning with a good kick-off. Fatty watched but
shook his head.
“Ah, leave him be, dad! He’s always tired,” Kak cried.
He sat down by his friend and was soon telling stories of their
southern travels. Kommana wanted to hear all about Jimmie
Muskrat, and Selby, and Noashak’s adventure with the deer. They
talked till nearly dark, and when the younger boy got back to the
tent he found his father and Okak in a friendly dispute concerning
the best route home.
Taptuna’s idea was to go westward, striking across the mouth of the
straits for Cape Baring, the southwest corner of Victoria Island,
where they would have a very good chance of killing a few polar
bears before the hardest frost set in, causing the open water to lie
farther and farther offshore, and leaving them to their regular life on
the ice catching seals. Okak as usual was raising objections. He still
had a quantity of trade goods, and things from their spring cache
made the load heavy. His neighbor pooh-poohed this, for they might
count on smooth going; but Okak was not to be easily moved. He
sat, brows knitted, a picture of worry, and tried to think up better
objections. Guninana glanced at him once or twice with a merry
twinkle in her eye. She knew his trouble—the poor chap was scared
stiff about bear hunting. The woman guessed right, but at that she
guessed only half his misery. Either way made Okak tremble in his
shoes. For days and days recollection of his cold ducking, with
renewed horror of snatching currents and bending ice, had been
haunting his memory. He did not forget it would be safer farther
west where the water flows more slowly—but what is the use of a
safety leading straight into the jaws of nasty, snarling bears? He
growled like a bear himself, seeing Taptuna wink at his wife.
In her heart of hearts Guninana sympathized with the nervous man.
She would have been better pleased to settle down on the ice
immediately, even if it meant eating seal and nothing but seal for
months; and so she was highly delighted when Okak suddenly burst
out:
“Two dogs are not sufficient! With only two men and two dogs the
results will be as poor as the hunting is risky, and all our time
wasted.”
Nobody answered this for it was sound reasoning. The little man sat
back rubbing his knees with a that-settles-it sort of superior manner.
“Alunak might join us,” Taptuna muttered, annoyed.
“He has promised his wife to go to Franklin Bay and try to meet the
Kabluna. She wants some steel needles.”
Guninana’s speech sounded gently satisfied; Okak observed it and
swelled with importance.
“Two dogs——” he began, intending to enlarge on his happy
inspiration, but it was just at this moment Kak entered.
“Who said ‘two dogs’?” the lively lad cried in a round, booming, out-
of-doors voice. “What about Kanik—my pup? I’d have you remember
we’ve got three dogs now!”
The resonant words shot like a boomerang through Okak’s self-
complacence. Instantly he knew the cause lost. He heard it in
Guninana’s little gasp; read it in his neighbor’s sparkling eyes bent
on the intruder.
“You think of everything, my boy. I had forgotten Kanik.”
Taptuna spoke quietly, but all saw his elation. He felt immensely
proud of Kak, and in that the boy’s mother must join him. Fresh
proof of her son’s cleverness always put Guninana into beaming
good humor; moreover, it is fun to play on the winning side. The
family joined forces against Okak and silenced his arguments if not
his fears.
They agreed to travel as far as Crocker River with Alunak’s party, and
this journey turned out harder and slower than anybody had
anticipated, for a strong wind from the northwest blew directly in
their faces all the way. At the river Okak made a final throw for
safety by trying to persuade their friends to join forces in bear
hunting at the eleventh hour. Alunak himself was minded to do so, if
it had not been for his wife’s fixed idea about needles. He had
promised, and the lady being a very dominant person meant to see
that he kept his promise. They all got into a great discussion over it,
which lasted while they were house-building and eating, and
commenced again the next morning. Nothing would turn the
woman; Guninana even offered to lend her a needle for as long as
they were in Victoria Island, but she held to her point. Perhaps she
was as curious to see the Kabluna as to inspect his trade goods; Kak
thought so anyway, and blazing with a wild hope suggested they
might all go on to Franklin Bay first. When his father answered “No,”
most emphatically, he grew tired of the merry argument and,
deciding to take his dog for a walk, went out alone.
Kanik leaped up, pawing his master’s shoulders, making no end of a
fuss and acting silly as a pup does; the pair were perfectly happy till
Sapsuk got on to what was afoot and whined, wagging his tail,
pleading to be allowed to go. In his present mood the boy thought
two a company and three a crowd, so he felt annoyed. Sapsuk might
be his favorite, but Kanik was his own—if you have ever possessed a
dog you will understand. Kak was so torn between the two that in
the end he took neither.
“You have to work hard, and it is better for you to rest,” he
admonished like a grandfather, and started off, his walk already half
spoiled. “If Sapsuk keeps this up I’ll never be able to teach the pup
anything!” the boy muttered fretfully, for the first time wishing his
friend had loved him a little less.
Conditions showed that the wind blowing against them all the way
along must have been here a heavy, continuous gale. It had piled
more ice into the western mouth of the straits than had ever been
known before. The coast rose high. From its cliffs Kak beheld great
masses of ice filling the whole expanse, rolling away billow on billow
like a prairie country, goodness knew how deep under the trackless,
gleaming snow.
“Jimminy!” thought the boy. “This old sea is going to take some
crossing!”
He questioned if Omialik had started and felt a pang considering
how near his hero might be at that minute and he unable to reach
him. Then recollection of Okak brought a grin. “Our neighbor wanted
it thick and he’s got it—perhaps he’ll be sorrier yet we didn’t travel
by the eastern straits. I wonder what the going really is like out
there?”
To think was to act with Kak. He immediately scrambled down the
cliffs and a half hour later was walking alone over the corrugated ice
field.
It was a shimmering sort of day. The sun struggled to penetrate the
clouds, but did not quite emerge. The world lay trackless, formless,
shadowless, a vast expanse of gray-white sky and gray-white snow.
This kind of light is far harder on the eyes than bright sunshine, and
since his snowblindness Kak had been very nervous about eyes. He
kept his screwed up, not looking intently at anything, nor paying
much attention to where he went, for he counted on the cliffs to
guide him back. He only wanted to get a general impression of what
their next march would be like and so strolled carelessly up a high
ridge for a better view.
All at once Kak felt himself falling. He instantly thrust out his elbows
so they would catch on the edges of the ice, for he knew what had
happened. Stepping heedlessly he had walked on to the snow roof
of a crevasse and had gone through into the crack. This is a
common form of Arctic accident. The boy expected to stop when he
had fallen as far at his waist, and to be able to hoist himself out,
none the worse for his adventure; but to his surprise and horror he
kept right on falling. The width of this chasm was so great that his
elbows did not reach the walls. For an instant Kak felt helplessly
angry—then the serious side broke on him. He was falling, falling—
where to? On what would he strike—ice or water? How far would he
fall? How hard would he strike? Sick with fear he tried to use his
frenzied wits. It darted into his mind like a javelin that they would
not know at home where he had gone, for snow so hard-driven by
the gale was trackless as a rock. How he wished now he had taken
either of the dogs, or both! He thought of Omialik, regretting
Herschel Island, and in the middle of his keenest sorrow for the
young marksman who would never be, both feet hit suddenly, smack
on glare ice, flew from under him, and pitched him shoulder on
against the solid wall. He slid down, smashing the back of his head,
and lay still. Pain mingled with relief. It seemed for a moment as if
nothing again could ever be so bad as that falling sensation. But the
brief happiness passed. He realized he was lying captive between
two high, hard, slippery sides, which towered above his head in
twilight to the snow roof of the crevasse, offering no way out of that
strange, cold prison. Above he could see the jagged hole he had torn
in falling, and beyond it the gray sky. Through a fresh tide crack in
the ice floor he saw water. Fear gripped him again when he thought
how a little less frost would have allowed him to go right splash into
it; for when an ice cake cracks it splits from top to bottom, leaving
open ocean. Had the storm which roofed the tunnel over brought a
spell of warm weather instead of cold, as storms often do, there
would have been no floor formed in the crevasse. Bad as his plight
was, things might have been infinitely worse. Suppose he had been
floundering and freezing now—drowning, down in the bottom of that
dismal jail without means of escape or alarm. Again, and this time in
a very different mood, he regretted leaving his faithful dog. Sapsuk
would have had sufficient intelligence to run and fetch Taptuna.
Kak knew very well nobody would come to help him, so he must
help himself. As a beginning he took stock of his condition. One hip
and shoulder were badly bruised and painful, and a goose egg was
already developing on his head; but no bones seemed to be broken,
nor could he find sprains or dislocations. So far so good. His first
idea was to cut steps in the face of the ice wall and climb out.
Putting his hand to his belt he found both knife and sheath had been
torn away. “Still, it must be here,” the boy said bravely, and
commenced looking around. The tide crack mocked him like an
open, laughing mouth. “Foxes! If it has gone in there!” he cried,
fumbling frantically under the snow which had showered down with
his fall. Presently his fingers rapped on a horn handle. He made one
grab and almost wept for joy. Just then his knife seemed his
salvation; but five minutes later it had lost half of its value. On trial
he found the sides were too far apart for him to support himself by a
braced arm or knee as he climbed, and walking straight up a
perpendicular, slippery surface by toe holes is an utter impossibility.
Kak now understood getting away was going to take all his invention
and nerve and strength. The first step was to learn his surroundings.
This crack might run smaller or lower at some other point. He set
out exploring. It was an eerie sort of business to turn his back on
the pool of light striking through the roof hole, and crawl over glare
ice, between those blue-white walls, into the very heart of the
stupendous jam he had so recently viewed with wonder from the
cliffs. On hands and knees the boy began his strange and thrilling
tour. His position brought him close to the floor, and once beyond
the showered snow he saw tracks in the hoarfrost on the ice which
made him flinch. He had company in the tunnel. The footsteps went
both ways, as if some poor trapped creature had run to and fro, to
and fro, in a crazy hunt for freedom. Kak knew very well what tracks
these were. Acute dread shuddered over him. “But the crevasse may
be long,” he comforted himself. “With luck I may get out before we
meet.”
He crawled for thirty yards, stood up, and tried to guess the height
of his prison. The snow roof looked thick and solid here, and though
some light filtered through it, and doubtless a little through the ice
itself, the gloom was sufficiently thick to confuse calculation. Space
seemed to yawn above him; Kak felt rather than saw those walls
were higher and wider apart; so he retreated to his first position
and, only waiting to take one long look up at the friendly sky, set out
in the opposite direction.
There was no question about it, the walls lowered toward this end.
Fired with hope our boy scuttled along like a crab. The ice lay
perfectly smooth, slippery as a ballroom floor. He crawled a few feet
and stopped to glance above, and crawled on, and stopped, till
familiarity made him careless. Very soon he was crawling and gazing
upward together, forgetful of everything but his anxiety to climb out.
Then suddenly the advanced arm plunged down splash into another
tide crack. Kak uttered a yap of surprise, snatched back his hand,
peeled off the wet mitt and dried his fingers quickly on his clothes. It
had not gone in above the wrist, but a wet mitt was going to be less
comfortable than a dry one; the captive felt vexed at his stupidity,
blamed his position for it and scrambling to his feet walked slowly,
steadying himself with his right arm against the wall, which bent at a
gentle angle. Soon he spied ahead a second pool of light, a second
scattering of snow from a hole in the ceiling. For an instant Kak felt
glad—misery loves company—then it dawned on him what had fallen
through, and his teeth chattered. This snow, packed and trodden
down, looked several days old. Would he find a dead thing here
entombed with him—or worse, a hungry living thing?
It took all the boy’s grit to make him go on. Only the sight of those
lowering sides lent him courage. His sole chance for safety might lie
hand in hand with this mysterious danger if the beast had elected to
live in the small end of the crack. Light was failing again as he
moved away from the second hole, and the darkness tortured his
trembling nerves. Cautiously the lad stole on. His right hand grasped
his knife, his left was ready for action; while he seemed to cling to
the slippery path by his toes.
On either hand the sides sloped downward. “If it keeps on like this
the crack will end in a cave,” Kak thought, “a cave with a top of soft
snow well within my reach.”
Sure enough! He came to another corner, rounded it timidly, and
found himself facing the end of the tunnel where the walls ran
sharply together, forming a narrow cave. In this cave, filling it
completely, stood a full-grown wolf. Its gaunt, gray form was partly
shrouded by gray gloom, but the yellow eyes looking out of that
triangular face were horribly alive. Kak stopped, choking back fear.
He swallowed. His breath caught and came in sobs, turn about. He
wanted to fly and was too frightened; so he just stood like a fool,
waiting for the famished animal to spring and devour him. The wolf
waited also.... Little by little, as nothing happened, the boy regained
his common sense. Of course the wolf would be scared, poor thing,
cornered that way with no means of escape! He saw it was petrified
by fear. It looked thin and hungry and was probably weak. Kak felt
very sorry for his fellow prisoner, yet he wanted to put distance
between them. One never knows the strength and wickedness of a
wild animal at bay.
The two stood regarding each other, neither of them moving. Kak
had the advantage—he could retreat. His brain worked madly.
“If I go back to the second hole,” he thought, “and try knocking
some more snow down and piling it up against the side of the
crevasse, possibly I can climb out there.”
Stealthily he edged away, keeping his eye on the foe till the curve of
the wall divided them; then he made tracks as fast as he could over
the glare ice.
Standing under the hole broken by the wolf’s fall Kak sent his knife
flying up against the roof; it fell back amid a tiny shower of snow.
He threw it again; a slightly heavier cloud descended. At each throw
a little more seemed to come down. The boy was all eagerness; he
tossed and tossed and tossed in a fury of excitement till he saw the
precious knife suddenly shoot up against the sky. For one terrifying
instant it looked as if it would fall outside on top of the crust. His
heart stopped beating. He shut his eyes. Hours seemed to pass
before the tinkle of copper on ice broke his tension.
“Bears and foxes! How could I have been so careless hopping about
that way and never giving a thought!”
Facing a large, ravenous wolf with a knife in one’s hand, and facing
the same beast unarmed are vastly different. This momentary shock
made it clear to Kak he was fairly well off, but it jarred his faith in
the new scheme. He was afraid now to throw with energy and
abandon, and the roof seemed too hardly packed to be broken by
half measures. He scraped the loose snow together with his feet,
piled it up, patting it hard by hand, stood on it and tried to reach the
top. But most of the mound had been lying on the ice floor and was
all powdery cold so that it broke under his weight.
“This will take days!” the boy cried in despair. “I’ll be hungry and
maybe freeze, or perhaps the others will give me up and go away.”
His fingers in the wet mitt felt bitterly cold. Taking it off he drew his
hand through the loose sleeve of his coat and shirt and cuddled it
against his warm body while he stood gazing at the height of those
forbidding sides. All the time his glance rested on their inaccessibility
his mind was busy reckoning how low they ran in the cave behind
the wolf.
“I’ve got to do it! I’ve got to do it! I must get out of here before
night,” wailed Kak. He turned and looked undaunted down the
tunnel.
“I’ve just got to!”
Screwing his courage to the breaking point and grasping his knife
more firmly the second prisoner crept forward to the angle in the
wall. He shoved his head around cautiously. There stood the wolf
exactly as Kak had left him. He seemed too frightened even to blink
his eyes.
Quite aside from the fear of combat Kak was reluctant to attack this
poor caged animal.
“If it only wasn’t so narrow there I could shove in and shove him out
—given a chance he’d split past me like the wind.”
But it was narrow in the cave, much too narrow for any maneuver of
that sort.
“I’ve got to kill him and haul him out! I haven’t any choice,” cried the
boy.

KAK RUSHED FORWARD WITH HIS KNIFE READY.

Kak rushed forward with his knife ready and his left arm thrown up
in front guarding his face. When the beast reared and hurled itself
for a grasp of the enemy’s throat its long jaws closed on the
shielding wrist. With a gasp of pain the boy flung his arm wide,
wrenching the wolf’s head clear around, and at the same second
stuck his blade deep into the side under its foreleg. Between the
double shock of the twist and the blow his victim lost its footing and
fell to the ground with a heavy crash, dragging the hunter down on
top of him. For a moment Kak rolled amid a convulsed mass of feet
and legs, then as the spasm ceased the vise grip on his arm relaxed,
and the animal fell limp. Such narrow quarters had offered no
chance for a fair fight; it was lunge, grab or be grabbed, and die.
The boy scrambled to his knees, withdrew his knife, dragged the
warm body out of the way, and with a shudder sprang from it into
the extreme end of the crevasse. For five minutes he worked off his
emotion by hacking snow like a madman. It fell around and over him
in showers, hiding the bloody trail that oozed across the ice and the
spatters from his wounded wrist of which, in his haste to get away,
he took no heed.
All at once the roof broke, came down like an avalanche, and the
fresh air streamed in. The boy stopped for a deep breath. He could
grasp the ice edge with his fingers, but it was still too high for him to
pull himself out. He worked swiftly, cutting blocks from the ceiling
and piling their fragments against the end of the crack; and all the
time it seemed as if that hideous wolf behind was rearing over him,
fixed-eyed and open-mouthed.
Kak was pretty tired and unstrung when finally he placed both hands
on the crusted snow and drew himself into freedom. How good the
air tasted. How heartening was the vast horizon sweep! He ran to
warm up, for it had been searchingly cold down in the bottom of
that deep ice pit. “Bhooo!” he shivered, nursing his sore arm.
Running soon set the healthy blood coursing in his veins; his body
tingled and his spirits rose.
As soon as his nerves grew normal Kak’s point of view changed. He
saw the hair-raising experience might be turned into splendid
adventure.
“Why not have some honor out of this?” the boy thought. So instead
of dashing home all trembling and excited, he held himself down to
a steady walk, stopped outside a minute to give old Sapsuk an
apologetic little love pat, also for the sake of seeming casual, and
then strode in.
“I’ve killed a wolf, dad,” he said. “It’s a thin, poor thing, but it will
help. See here.” And he threw his bloody knife on the floor by way of
evidence.
Guninana wasted no time on the weapon; one glance at his sleeve
and mitt set her bustling around for rude means of relief. The others
cried out in amazement, examined the knife, bombarded him with
questions, laughed and clapped like children, quaked and marveled,
while Kak wallowed in praise and the show of his mother’s
attentions. Okak was for going after the carcass at once; but the
hunter assured him the meat was safely cached, and burst into
laughter at what he called a good joke—then he had to explain.
Unable any longer to keep up his hero pose he told the whole story.
It was an amazing story. Such ice formations are more common in
the Antarctic than the north. Everybody flocked over to see the
crevasse and help bring the victim home. Taptuna skinned the wolf
beautifully; and you may be sure the boy was very careful to pack
his trophy next morning, when the parties separated, each going its
own way with perfect understanding, and much calling of gay good-
bys back and fore.
Our friends were in high spirits. No one really minded the difficulties
of rolling ridges and heavy travel. Guninana gloried in her son; Kak
was triumphant; Taptuna seemed as proud of his new sled as
Noashak of her coat with red trimmings. And Okak had enough trade
goods to make him a well-to-do man.
Their summer trip had prospered through strenuous labor and
thrilling feats, and they all looked forward to their winter on the ice
as a well-earned holiday.
 TRANSCRIBER’S NOTES:
Obvious typographical errors have been corrected.
Inconsistencies in hyphenation have been
standardized.
The illustration listed as facing page 43 in the List of
Illustrations does not appear in the original book.
The caption for the Frontispiece was added by the
transcriber based upon the List of Illustrations.
*** END OF THE PROJECT GUTENBERG EBOOK KAK, THE COPPER
ESKIMO ***

Updated editions will replace the previous one—the old editions

will be renamed.

Creating the works from print editions not protected by U.S.

copyright law means that no one owns a United States
copyright in these works, so the Foundation (and you!) can copy
and distribute it in the United States without permission and
without paying copyright royalties. Special rules, set forth in the
General Terms of Use part of this license, apply to copying and
distributing Project Gutenberg™ electronic works to protect the
PROJECT GUTENBERG™ concept and trademark. Project
Gutenberg is a registered trademark, and may not be used if
you charge for an eBook, except by following the terms of the
trademark license, including paying royalties for use of the
Project Gutenberg trademark. If you do not charge anything for
copies of this eBook, complying with the trademark license is
very easy. You may use this eBook for nearly any purpose such
as creation of derivative works, reports, performances and
research. Project Gutenberg eBooks may be modified and
printed and given away—you may do practically ANYTHING in
the United States with eBooks not protected by U.S. copyright
law. Redistribution is subject to the trademark license, especially
commercial redistribution.

START: FULL LICENSE

THE FULL PROJECT GUTENBERG LICENSE