Fuzzy Computing in Data Science E-Book

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: Smart and Sustainable Intelligent Systems
• Sprache: Englisch

Seitenzahl: 429

Veröffentlichungsjahr: 2022

Beschreibung

FUZZY COMPUTING IN DATA SCIENCE This book comprehensively explains how to use various fuzzy-based models to solve real-time industrial challenges. The book provides information about fundamental aspects of the field and explores the myriad applications of fuzzy logic techniques and methods. It presents basic conceptual considerations and case studies of applications of fuzzy computation. It covers the fundamental concepts and techniques for system modeling, information processing, intelligent system design, decision analysis, statistical analysis, pattern recognition, automated learning, system control, and identification. The book also discusses the combination of fuzzy computation techniques with other computational intelligence approaches such as neural and evolutionary computation. Audience Researchers and students in computer science, artificial intelligence, machine learning, big data analytics, and information and communication technology.

Leseprobe

Table of Contents

Cover

Series Page

Title Page

Copyright Page

Dedication Page

Preface

Acknowledgement

1 Band Reduction of HSI Segmentation Using FCM

1.1 Introduction

1.2 Existing Method

1.3 Proposed Method

1.4 Experimental Results

1.5 Analysis of Results

1.6 Conclusions

References

2 A Fuzzy Approach to Face Mask Detection

2.1 Introduction

2.2 Existing Work

2.3 The Proposed Framework

2.4 Set-Up and Libraries Used

2.5 Implementation

2.6 Results and Analysis

2.7 Conclusion and Future Work

References

3 Application of Fuzzy Logic to Healthcare Industry

3.1 Introduction

3.2 Background

3.3 Fuzzy Logic

3.4 Fuzzy Logic in Healthcare

3.5 Conclusions

References

4 A Bibliometric Approach and Systematic Exploration of Global Research Activity on Fuzzy Logic in Scopus Database

4.1 Introduction

4.2 Data Extraction and Interpretation

4.3 Results and Discussion

4.4 Bibliographic Coupling of Documents, Sources, Authors, and Countries

4.5 Conclusion

References

5 Fuzzy Decision Making in Predictive Analytics and Resource Scheduling

5.1 Introduction

5.2 History of Fuzzy Logic and Its Applications

5.3 Approximate Reasoning

5.4 Fuzzy Sets vs Classical Sets

5.5 Fuzzy Inference System

5.6 Fuzzy Decision Trees

5.7 Fuzzy Logic as Applied to Resource Scheduling in a Cloud Environment

5.8 Conclusion

References

6 Application of Fuzzy Logic and Machine Learning Concept in Sales Data Forecasting Decision Analytics Using ARIMA Model

6.1 Introduction

6.2 Model Study

6.3 Methodology

6.4 Result Analysis

6.5 Conclusion

References

7 Modified m-Polar Fuzzy Set ELECTRE-I Approach

7.1 Introduction

7.2 Implementation of m-Polar Fuzzy ELECTRE-I Integrated Shannon’s Entropy Weight Calculations

7.3 Application to Industrial Problems

7.4 Conclusions

References

8 Fuzzy Decision Making: Concept and Models

8.1 Introduction

8.2 Classical Set

8.3 Fuzzy Set

8.4 Properties of Fuzzy Set

8.5 Types of Decision Making

8.6 Methods of Multiattribute Decision Making (MADM)

8.7 Applications of Fuzzy Logic

8.8 Conclusion

References

9 Use of Fuzzy Logic for Psychological Support to Migrant Workers of Southern Odisha (India)

9.1 Introduction

9.2 Objectives and Methodology

9.3 Effect of COVID-19 on the Psychology and Emotion of Repatriated Migrants

9.4 Findings

9.5 Way Out for Strengthening the Psychological Strength of the Migrant Workers through Technological Aid

9.6 Conclusion

References

10 Fuzzy-Based Edge AI Approach: Smart Transformation of Healthcare for a Better Tomorrow

10.1 Significance of Machine Learning in Healthcare

10.2 Cloud-Based Artificial Intelligent Secure Models

10.3 Applications and Usage of Machine Learning in Healthcare

10.4 Edge AI: For Smart Transformation of Healthcare

10.5 Edge AI-Modernizing Human Machine Interface

10.6 Significance of Fuzzy in Healthcare

10.7 Conclusion and Discussions

References

11 Video Conferencing (VC) Software Selection Using Fuzzy TOPSIS

11.1 Introduction

11.2 Video Conferencing Software and Its Major Features

11.3 Fuzzy TOPSIS

11.4 Sample Numerical Illustration

11.5 Conclusions

References

12 Estimation of Nonperforming Assets of Indian Commercial Banks Using Fuzzy AHP and Goal Programming

12.1 Introduction

12.2 Research Model

12.3 Result and Discussion

12.4 Conclusion

References

13 Evaluation of Ergonomic Design for the Visual Display Terminal Operator at Static Work Under FMCDM Environment

13.1 Introduction

13.2 Proposed Algorithm

13.3 An Illustrative Example on Ergonomic Design Evaluation

13.4 Conclusions

References

14 Optimization of Energy Generated from Ocean Wave Energy Using Fuzzy Logic

14.1 Introduction

14.2 Control Approach in Wave Energy Systems

14.3 Related Work

14.4 Mathematical Modeling for Energy Conversion from Ocean Waves

14.5 Proposed Methodology

14.6 Conclusion

References

15 The m-Polar Fuzzy TOPSIS Method for NTM Selection

15.1 Introduction

15.2 Literature Review

15.3 Methodology

15.4 Case Study

15.5 Results and Discussions

15.6 Conclusions and Future Scope

References

16 Comparative Analysis on Material Handling Device Selection Using Hybrid FMCDM Methodology

16.1 Introduction

16.2 MCDM Techniques

16.3 The Proposed Hybrid and Super Hybrid FMCDM Approaches

16.4 Illustrative Example

16.5 Results and Discussions

16.6 Conclusions

References

17 Fuzzy MCDM on CCPM for Decision Making: A Case Study

17.1 Introduction

17.2 Literature Review

17.3 Objective of Research

17.4 Cluster Analysis

17.5 Clustering

17.6 Methodology

17.7 TOPSIS Method

17.8 Fuzzy TOPSIS Method

17.9 Conclusion

17.10 Scope of Future Study

References

Index

Also of Interest

End User License Agreement

List of Tables

Chapter 1

Table 1.1 Pixels clustered based on PSC (EEOC) with K-means and FCM for Sali...

Table 1.2 Pixels clustered based on PSC (EEOC) with K-means and FCM for Indi...

Table 1.3 Pixels clustered based on PSC (EEOC) with K-means and FCM for Sali...

Table 1.4 Pixels clustered based on PSC (EEOC) with K-means and FCM for Pavi...

Table 1.5 Pixels clustered based on PSC (EEOC) with K-means and FCM for Pavi...

Table 1.6 Elapsed time in seconds for PSC (EEOC) with K-means and FCM.

Table 1.7 Fitness value for PSC (EEOC) with K-means and FCM.

Chapter 4

Table 4.1 Prominent affiliations contributing toward fuzzy analysis.

Table 4.2 Top Journals publishing fuzzy-related works.

Table 4.3 Major contributing countries toward fuzzy research articles.

Table 4.4 Renowned authors contributing toward fuzzy analysis.

Chapter 5

Table 5.1 Tasks with parameters.

Table 5.2 Assigning fuzzy membership values.

Chapter 6

Table 6.1 Seed types.

Table 6.2 Location-based seed sales.

Table 6.3 Season-based seed data.

Chapter 7

Table 7.1 Methods applied for cutting fluid selection.

Table 7.2 Decision matrix for neat oil selection problem [3].

Table 7.3 Normalized decision matrix for the neat oil selection problem.

Table 7.4 Weight calculated by Shannon’s entropy method.

Table 7.5 Weight multiplied matrix for cutting fluid selection problem.

Table 7.6 Concordance set.

Table 7.7 Discordance set.

Table 7.8 Comparison between modified similarity and TOPSIS methods for neat...

Table 7.9 Outranking relationship between alternatives.

Table 7.10 Methods for FMS selection.

Table 7.11 Decision matrix for FMS selection problem [27].

Table 7.12 Normalized decision matrix for the FMS selection problem.

Table 7.13 Shannon’s entropy weight for variables

Table 7.14 Weight multiplied matrix for FMS selection problem.

Table 7.15 Concordance set.

Table 7.16 Discordance set.

Table 7.17 Comparison between EVAMIX, COPRAS, and m-polar fuzzy ELECTRE-I me...

Table 7.18 Outranking relationship between alternatives.

Chapter 8

Table 8.1 Variables for individual decision making of job selection

Table 8.2 Creating a decision matrix.

Chapter 9

Table 9.1 Variables considered for psychological and emotional effect of the...

Chapter 11

Table 11.1 Fuzzy linguistic ratings.

Table 11.2 Linguistic scale for alternative rankings.

Table 11.3 Importance weight of each criterion.

Table 11.4 Linguistic rating variable for evaluate the rating of alternative...

Table 11.5 The fuzzy decision matrix and fuzzy weights of alternatives.

Table 11.6 Fuzzy normalized decision matrix.

Table 11.7 Fuzzy weighted normalized decision matrix.

Table 11.8 Distance calculation.

Table 11.9 Closeness coefficient.

Chapter 12

Table 12.1 Gross NPA in priority and non-priority sectors.

Table 12.2 Interdependency of different economic sectors.

Table 12.3 Original Saaty’s scale for pairwise comparison.

Table 12.4 Extension of Saaty’s scale.

Table 12.5 Fuzzified matrix.

Table 12.6 Average growth rate in different sectors.

Table 12.7 Arrangement of priority and non-priority sectors as per weight.

Table 12.8 Priority arrangement.

Table 12.9 Comparison of gross NPA in SBI annual report 2020–2021 to gross N...

Chapter 13

Table 13.1 Five degrees of linguistic variables for assessing subjective att...

Table 13.2 Linguistic variables, abbreviation and TFN for weights.

Table 13.3 Decision matrix in linguistic variables assessed by user (expert)...

Table 13.4 Decision matrix in terms of linguistic variable estimated by user...

Table 13.5 Decision matrix in terms of linguistic variable estimated by user...

Table 13.6 Weights in linguistic variables estimated by users.

Table 13.7 Performance contribution of individual criterion.

Table 13.8 Performance index and ranking order of alternatives.

Chapter 14

Table 14.1 Fuzzy rules for optimization of parameters.

Chapter 15

Table 15.1 Comparison of different NTM process selection methods [11].

Table 15.2 Decision matrix (DM), Example from [5].

Table 15.3 Normalized DM with linear max N1 method [16].

Table 15.4 Pairwise comparison matrix for criteria.

Table 15.5 The rank of criteria based on weight their weights.

Table 15.6 Weight multiplied matrix.

Table 15.7 Euclidean distance of alternatives from a positive ideal solution...

Table 15.8 The rank of NTM processes using the m-polar fuzzy TOPSIS algorith...

Table 15.9 Decision matrix, example from [5].

Table 15.10 Normalized decision matrix with linear max N1 method.

Table 15.11 Weight multiplied matrix.

Table 15.12 Euclidean distance of alternatives from a positive ideal solutio...

Table 15.13 The rank of alternatives based on RCC.

Table 15.14 Result validation for m-polar fuzzy TOPSIS methodology with prev...

Table 15.15 Data input uncertainty margin for a positive ideal solution with...

Table 15.16 Data input uncertainty margin for a positive ideal solution with...

Chapter 16

Table 16.1 Fuzzy numbers, linguistic variables, and scale of TFNs for rating...

Table 16.2 Fuzzy numbers, linguistic variables, TFN, (TFN)

-1

for criteria we...

Table 16.3 Decision matrix for MHD selection problem.

Table 16.4 Decision matrix in terms of triangular fuzzy numbers.

Table 16.5 Weighted normalized performance ratings.

Table 16.6 MHD selection indices and ranking by proposed super hybrid approa...

Table 16.7 Spearman rank correlation coefficients.

Chapter 17

Table 17.1 Weighted normalized data.

Table 17.2 Grouping of policies into different clusters.

Table 17.3 Selection of alternative.

Table 17.4 (a) Ideal crop plan from the planning model.

Table 17.4 (b) Pay-off matrix.

Table 17.5 Fuzzy decision matrix.

Table 17.6 Normalized decision matrix.

Table 17.7 Weighted normalized decision matrix.

Table 17.8 Separation measures.

Table 17.9 The relative closeness to ideal solution.

Table 17.10 Interval performance measure.

Table 17.11 Crisp performance measure.

Table 17.12 Rank of different policies by fuzzy TOPSIS method.

List of Illustrations

Chapter 1

Figure 1.1 Flowchart for hyperspectral image segmentation using EEOC.

Figure 1.2 Band reduction using K-means algorithm.

Figure 1.3 Band reduction using Fuzzy C-means.

Figure 1.4 DB Index Graph. (a) Salina_A scene (b) Indian Pines (c) Salinas V...

Figure 1.5 Results for Salinas_A, (a) K-means + PSC (EEOC) (b) FCM + PSC (EE...

Figure 1.6 Results for Indian Pines (a) K-means + PSC (EEOC) (b) FCM +PSC (E...

Figure 1.7 Results for Salinas Valley. (a) K-means + PSC (EEOC) (b) FCM + PS...

Figure 1.8 Results for Pavia University. (a) K-means + PSC (EEOC) (b) FCM + ...

Figure 1.9 Results for Pavia Centre. (a) K-means + PSC (EEOC) (b) FCM + PSC ...

Chapter 2

Figure 2.1 The proposed framework.

Figure 2.2 Some of the libraries used.

Figure 2.3 The dataset and directory.

Figure 2.4 Training the model.

Figure 2.5 Training the model.

Figure 2.6 Evaluation of the networks and the precision scores.

Figure 2.7 Training loss and accuracy graph on the COVID-19 dataset.

Figure 2.8 In real time with cap, face detected without a mask with 100% acc...

Figure 2.9 In real time, with cap, face detected with the mask with 99% accu...

Figure 2.10 In real time, the persons face in the image is detected without ...

Figure 2.11 In real time the persons face in the image is detected with the ...

Figure 2.12 In real time, the persons face in the image is detected without ...

Figure 2.13 In real time ,the persons face in the image is detected with the...

Chapter 3

Figure 3.1 The basic components of FL.

Figure 3.2 The advantages of a Fuzzy Logic system.

Figure 3.3 The disadvantages of Fuzzy Logic system.

Figure 3.4 Fuzzy Logic in healthcare.

Figure 3.5 Lists of FL applications in healthcare industries.

Chapter 4

Figure 4.1 Stages of Bibliometric analysis for fuzzy logic.

Figure 4.2 Yearwise publication and citation count of fuzzy logic.

Figure 4.3 Major subject areas using fuzzy logic.

Figure 4.4 Overlay visualization of countries.

Figure 4.5 Network visualization of the co-authorship of authors contributin...

Figure 4.6 Cocitation analysis of cited authors.

Figure 4.7 Network visualization of co-occurrence of author keywords.

Figure 4.8 Overlay visualization of bibliographic coupling of documents.

Figure 4.9 Overlay visualization of bibliographic coupling of sources.

Figure 4.10 Overlay visualization of bibliographic coupling of authors.

Figure 4.11 Overlay visualization of bibliographic coupling of countries.

Chapter 5

Figure 5.1 Difference between the conventional set theory and fuzzy logic.

Figure 5.2 Comparison of the membership function of fuzzy sets with classica...

Figure 5.3 Fuzzy inference system.

Figure 5.4 Showing partially constructed fuzzy decision tree.

Chapter 6

Figure 6.1 Sample seed dataset.

Figure 6.2 A four-phase business cycle.

Figure 6.3 Seed sales data.

Figure 6.4 Location-based seed data.

Figure 6.5 Spinach sales.

Figure 6.6 Gingelly sales.

Figure 6.7 Black gram sales.

Figure 6.8 Fodder sorghum sales.

Figure 6.9 Prediction for black gram.

Figure 6.10 Prediction for spinach dataset.

Figure 6.11 Prediction for gingelly dataset.

Figure 6.12 Prediction for fodder dataset.

Chapter 7

Figure 7.1 Directed graph for cutting fluid selection.

Figure 7.2 Directed graph for FMS selection.

Chapter 8

Figure 8.1 Comparison of Boolean logic with fuzzy logic.

Chapter 10

Figure 10.1 Applications and usage of machine learning in healthcare.

Figure 10.2 Role of edge in healthcare.

Figure 10.3 The approach of multiattribute decision model.

Chapter 12

Figure 12.1 Intersection between two triangular fuzzy numbers (M

1

& M

2

).

Figure 12.2 Feasibility result of goal programming model.

Chapter 13

Figure 13.1 Performance index of alternatives.

Figure 13.2 Ranking order of the alternatives.

Chapter 14

Figure 14.1 Types of buoys for ocean wave.

Figure 14.2 Waves motion and depth.

Figure 14.3 Block diagram of a sea wave energy harvesting device.

Figure 14.4 Conceptual model wave energy generation.

Figure 14.5 Proposed wave optimization model.

Chapter 15

Figure 15.1 Graphical presentation of the rank performance of alternatives f...

Figure 15.2 Graphical presentation of the rank performance of alternatives f...

Chapter 16

Figure 16.1 Decision making framework for MHD selection.

Figure 16.2 Comparison of ranking order.

Figure 16.3 Ranks of MHDs by super hybrid approach.

Guide

Cover Page

Series Page

Title Page

Copyright Page

Dedication Page

Preface

Acknowledgement

Table of Contents

Begin Reading

Index

Also of Interest

Wiley End User License Agreement

Pages

ii

iii

iv

v

xvii

xviii

xix

xxi

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

79

80

81

82

83

84

85

86

87

88

89

90

91

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

253

254

255

256

257

258

259

260

261

262

263

264

265

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

329

330

331

332

333

334

335

336

337

Scrivener Publishing100 Cummings Center, Suite 541JBeverly, MA 01915-6106

Sustainable Computing and Optimization

Series Editor: Prasenjit Chatterjee, Morteza Yazdani and Dilbagh Panchal

Scope: The objective of “Sustainable Computing and Optimization” series is to bring together the global research scholars, experts, and scientists in the research areas of sustainable computing and optimization from all over the world to share their knowledge and experiences on current research achievements in these fields. The series aims to provide a golden opportunity for global research community to share their novel research results, findings, and innovations to a wide range of readers, present globally. Data is everywhere and continuing to grow massively, which has created a huge demand for qualified experts who can uncover valuable insights from data. The series will promote sustainable computing and optimization methodologies in order to solve real life problems mainly from engineering and management systems domains. The series will mainly focus on the real life problems, which can suitably be handled through these paradigms.

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

Fuzzy Computing in Data Science

Applications and Challenges

Edited by

Sachi Nandan MohantyPrasenjit Chatterjee

and

Bui Thanh Hung

This edition first published 2023 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© 2023 Scrivener Publishing LLCFor 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 otherwise, 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 Headquarters111 River Street, Hoboken, NJ 07030, USA

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

Limit of Liability/Disclaimer of WarrantyWhile the publisher and authors have used their best efforts in preparing this work, they make no representations 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 representatives, 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 information does not mean that the publisher and authors endorse the information or services the organization, 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 978-1-119-86492-9

Cover image: Pixabay.ComCover design by Russell Richardson

Dedication

The Editors would like to dedicate this book to their parents, life partners, children, students, scholars, friends and colleagues.

Preface

The term fuzzy logic describes the concept of incomplete information. In real life, decision makers frequently encounter situations when it becomes very tough to apprehend whether a statement is true or false. Fuzzy logic allows a great deal of flexibility in reasoning at the time. The fuzzy algorithms aid in the solution of a problem and make the best feasible decision after taking into account all relevant data. They are mostly used in rule-based automatic controllers and based on the likelihood of an individual output being assigned to a particular input. Several organizations use advanced data analysis and business intelligence models to boost their performance. These analytical models entail working with data from a variety of sources. Fuzzy algorithms are an example of the transition from accuracy to imprecision. The present book aims to present how fuzzy computing can be explored to handle several challenges in real time data science applications. Machine learning approaches are used in computing environments to talk about intelligence, predictive and prescriptive features in understanding data analytics and the ways to deal them respectively. Fuzzy based systems are targeted to prophesy recommended improvements for better decision making in several domains and fields. Various robust traits of fuzzy systems paves its applications in Information Technology, Tourism, Education, E-Commerce, Agriculture, Health Care, Manufacturing, Sales, Retail and many more.

This book explains how to use various fuzzy-based models to solve real-time industrial challenges. The book is organized in 17 chapters. The inclusion of extensive guidance for using fuzzy-based models in each chapter is a unique feature of this book. The following is a chapter-by-chapter summary:

Chapter 1 discusses about efficient feature reduction and extraction with efficient search strategies to extract a set of spectral bands from a given imagery. In Chapter 2, a fuzzy-based framework is designed to detect and warn if a user is wearing a mask correctly or not when in public locations, using public services, or even while conversing with someone. This framework can be used in public places to monitor the public and alert users if they are not wearing their mask appropriately. Chapter 3 examines a fundamental FL framework and its applicability in the healthcare industry. A descriptive study was undertaken to look into and focus on the ranking, classification, data mining, feature selection, pattern recognition, and optimization processes in healthcare decision-making. Chapter 4 provides a structural overview and systematic review of the extensive research work that has been done on the use of fuzzy logic in various dimensions, as well as to assist researchers in gaining understanding into the issue. Chapter 5 presents a fuzzy decision making application for resource scheduling problems. Chapter 6 focuses on sales data analysis using four different types of seeds: spinach, black gramme, ginseng, and fodder sorghum. The proposed work used a fuzzy logic technique called clustering and artificial intelligence algorithms to evaluate sales of various seeds over time, as well as the performance of sales in a specific location. In Chapter 7, a single-pole (single-valued) decision making problem is solved using Shannon’s entropy weight calculation integrated m-polar fuzzy ELECTRE-I methodology. For demonstration purposes, two industrial selection problems are considered: flexible production system selection and cutting fluid selection. The concept of a fuzzy set in decision-making is discussed in Chapter 8. This chapter provides an outline of how fuzzy set theory can be used to various decision-making issues. The purpose of Chapter 9 is to describe the psychological state of migrant workers in Covid-19. Migrant workers from Odisha’s two southern districts (Khurdha and Gajapati) have been considered. In addition, this chapter discusses how fuzzy logic might aid in improving the psychological well-being of such migrants. Chapter 10 describes a fuzzy-based edge AI method for smart healthcare transformation for a better future. Chapter 11 attempts to provide a complete guideline for picking the best video-conferencing software for a higher education institution based on their needs, and then uses the Fuzzy-Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method to rank the various alternatives. Chapter 12 presents an application of Fuzzy Analytic Hierarchy Process (AHP) with extended Saaty’s scale and Goal Programming model for a case study of Indian commercial bank. Chapter 13 investigates a new fuzzy-based mathematical approach for analysing and evaluating static workstation performance parameters in order to make accurate and correct decisions. In Chapter 14, a model for calculating the power created by ocean waves is described. Furthermore, by using a fuzzy optimizer, it can maximise the overall power generated. In Chapter 15, the m-polar fuzzy set TOPSIS approach is used to answer the challenge of selecting a non-traditional machining process (NTM). Chapter 16 attempts to select the optimal material handling devices by combining a number of traditional fuzzy group decision-making models. In Chapter 17, a theoretical study of critical chain project management (CCPM) is offered, and the practical application of CCPM is demonstrated.

The EditorsAugust 2022

Acknowledgement

The editors wish to express their warm thanks and deep appreciation to those who provided input, support, constructive suggestions, and comments and assisted in editing and proofreading of this book.

This book would not have been possible without the valuable scholarly contributions of authors across the globe.

The editors avow the endless support and motivation from their family members and friends.

The editors are very grateful to all the members who served in the editorial and review process of the book.

Mere words cannot express the editors’ deep gratitude to the entire Scrivener Publishing team, particularly Mr. Martin Scrivener for keeping faith and showing the right path to accomplish this very timely book.

Finally, the editors use this opportunity to thank all the readers and expect that this book will continue to inspire and guide them for their future endeavour.

The Editors

1Band Reduction of HSI Segmentation Using FCM

V. Saravana Kumar1*, S. Anantha Sivaprakasam2, E.R. Naganathan3, Sunil Bhutada1 and M. Kavitha4

1 Department of IT, SreeNidhi Institute of Science and Technology, Hyderabad, India

2 Department of CSE, Rajalakshmi Engineering College, Chennai, India

3 Department of CSE, Koneru Lakshmaiah Education Foundation, Vaddeswaram, AP, India

4 Research Scholar, Manonmaniam Sundaranar University, Tirunelveli, India

Abstract

Hyperspectral has carried hundreds of nonoverlapping spectral channels of a specified scene, clustering is one of the approaches for diminishing the size of these large data sets. Segmentation is intricate for the raw data; however, it is likely for the reduced band of HSI. To lessen the band size, the classical clustering methods for example K-means, Fuzzy C-means are accomplished. An integrated image segmentation procedure built on interband clustering and intraband clustering is proposed. The interband clustering is performed by K-means clustering and Fuzzy C-means clustering algorithms, despite the fact the intraband clustering is executed using particle swarm segmentation (PSO) clustering algorithm. The performance of the K-means algorithm is subject to initial cluster centers. Besides, the final partition should be contingent on the initial configuration. The clustering consequences have profoundly been subject to the number of clusters stated. It is essential to provide refined direction for defining the number of clusters with the purpose of attaining appropriate clustering consequences. Davies Bouldin (DB) index is one of the reliable methods to outline the number of clusters for these clustering methods. The hyperspectral bands are clustered, and a band which has extreme variance from each cluster is preferred. This tactic forms the diminished set of bands. PSC (EEOC) accomplishes the segmentation process on the reduced bands. In conclusion, there is a comparison of the result produced for K-means worked with EEOC and FCM worked with EEOC in various HSI scenes.

Keywords: K-means, Fuzzy C-means, band reduction, PSO, cluster, centroid

1.1 Introduction

Hyperspectral [14] has carried hundreds of nonoverlapping spectral channels [11] of a given scene, clustering [13, 21] is one of the methods for reducing the size of these large data sets. Despite displaying the size of hyperspectral scene [17], the device does not support to display the scene directly. Segmentation is complicated for the raw data, whereas it is possible for the reduced band scene. Even though hyperspectral data [3] can provide finely resolved details about the spectral properties [4] to be identified, it also has some limitation. When dealing with such high-dimensional data [24, 25], one is faced with the “curse of dimensionality” problem One popular way to tackle the curse of dimensionality [7] is to employ a feature extraction technique [22]. To diminish the band size, the classical clustering methods [30], such as K-means, Fuzzy C-means [27] are handled. The former can be sensitive to the initial centers, while the results from the latter depend on the initial weights. Here, an integrated image segmentation [2] process based on interband clustering and intraband clustering is proposed. The interband clustering is performed by K-means clustering or Fuzzy C-means clustering algorithms, whereas the intraband clustering is executed using particle swarm segmentation (PSO) clustering algorithm. The performance of the K-means algorithm depends on initial cluster centers. Besides, the final partition depends on the initial configuration. The clustering results have heavily depended on the number of clusters specified. It is necessary to provide educated guidance for determining the number of clusters in order to achieve appropriate clustering results. Davies Bouldin (DB) index is one of the reliable methods to determine the number of clusters for these clustering methods. The hyperspectral bands [18] are clustered and a band [34], which has maximum variance, from each cluster is chosen. This forms the reduced set of bands. PSC (EEOC) [29] performs the segmentation process on the reduced bands. Finally, the result produced from K-means worked with EEOC and FCM worked with EEOC in various HSI [19] scenes was compared.

1.2 Existing Method

1.2.1 K-Means Clustering Method

Theoretically, K-means [1] is a typical algorithm. Now that it is elementary and expeditious, it is attractive in practice. To begin with, it segregates the input dataset into K-lusters. Each cluster is described by an adaptively changing centroid, starting from some initial values named seed points. K-means enumerates the squared distances between the inputs and centroids, and assigns inputs to the nearest centroid. The procedure continues until there is no significant change in the location of class mean vectors between successive iteration of the algorithms. Apparently, the performance of the K-means algorithm depends on initial cluster centers, whereas the final partition depends on the initial configuration.

1.2.2 Fuzzy C-Means

In Fuzzy C-means clustering [8], data elements can belong to more than one cluster and associated with each element is a set of membership levels. FCM clustering [9] is a process of assigning these membership levels, and then using them to assign data elements to one or more clusters. The aim of FCM [10] is to minimize an objective function. Contrary to traditional clustering analysis methods, which distribute each object to a unique group, fuzzy clustering algorithms gain membership values between 0 and 1 that indicate the degree of membership for each object to each group. Obviously, the sum of the membership values for each object to all the groups is definitely equal to 1. Different membership values show the probability of each object to different groups.

The main limitation of the FCM algorithm [16] is its sensitivity to noise. The FCM algorithm implements the clustering task for a data set by minimizing an objective-function subject to the probabilistic constraint that the summation of all the membership degrees of every data point to all clusters must be one. This constraint results in the problem of this membership assignment, that noise is treated the same as points close to the cluster centers. However, in reality, these points should be assigned very low or even zero membership in either cluster.

Like K-means, the clustering results have heavily depended on the number specified. It is also necessary to provide an educated guidance for determining the number cluster in order to achieve appropriate clustering results. Davies Bouldin (DB) index is one of the reliable methods to determine the number of clusters for these clustering methods.

1.2.3 Davies Bouldin Index

Davies Bouldin index was introduced in 1979 by David L. Davies and Donald W. Bouldin. It is one of the methods for evaluating clustering algorithms. This is an internal evaluation scheme, where the validation of how well the clustering has done using quantities and features inherent to the dataset.

Many other distance metrics can be used, in the case of manifolds and higher dimensional data, where the Euclidean distance may not be the best measure for determining the clusters. It is important to note that this distance metric has to match with the metric used in the clustering scheme itself for meaningful results.

Davies-Bouldin (DB) index is dependent both on the data, as well as the algorithm. The Davies–Bouldin index measures the average of similarity between each cluster and its most similar one. As the clusters have to be compact and separated the lower Davies-Bouldin index produces a better cluster configuration.

In this work, the interband clustering and intraband clustering approach has proposed. In interband clustering part; existing method such as K-means and Fuzzy C-means method are applied to reduce the band size. In intraband clustering part; a lightweight algorithm, namely Enhanced Estimation of Centroid (EEOC) [33] is proposed. This method is examined with the abovementioned clustering method by applying in various hyperspectral scenes.

1.2.4 Data Set Description of HSI

The dataset contains a variety of hyperspectral remote sensing [15], which are acquired from airborne and satellite. In this work, certain data, such as Salinas A, Salinas Valley, Indian Pines, Pavia University, and Pavia Centre, are handled.

The original scene and its corresponding ground truth image are downloaded from the link. http://www.ehu.eus/ccwintco/index.p...Hyperspectral_Remote_Sensing_Scenes

These scenes are a widely used benchmark for testing the accuracy of hyperspectral data [20] classification [23, 26] and segmentation [31].

1.3 Proposed Method

1.3.1 Hyperspectral Image Segmentation Using Enhanced Estimation of Centroid

These topics explain the integration of intraband and interband cluster approach for segmentation of hyperspectral image in a synergistic fashion. It demonstrates the advantage of the advanced properties of both analysis techniques in combined fashion of clustering method. The ultimate goal is to improve the analysis and interpretation of hyperspectral image. Owing that, classifications [5, 12], as well as segmentation [6], face problems related to the extremely high dimensionality [32] of the hyperspectral images are dominated by mixed pixels and un-mixing techniques are crucial for a correct interpretation and exploitation of the data. This chapter explored the integration of intraband and interband clustering methods for segmentation process with the ultimate goal of obtaining more accurate methods for the analysis of hyperspectral scene without increasing significantly the computational complexity of the process. There are many things to do for achieving this goal; to begin with, to improve unsupervised segmentation [28] by learning a task-relevant measure of spectral similarity from the feature matrix approaches. Besides, interband and intraband cluster techniques for segmentation are proposed. In addition, various existing clustering techniques are applied to interband part. Moreover, for improving the accuracy of various clustering methods, Davies-Bouldin Index is used to determine the number of clusters.

Furthermore, for intraband cluster, a new and novel algorithm entitled as Enhanced Estimation of Centroid (EEOC), which is the modification of the particle swarm clustering method, is proposed. The modification is that the particle has updated their positions and computed the distance matrix only once per iteration. In addition, performances about these methods are evaluated in different scenarios. Performance measurements are used in this research to investigate the efficacy of this system.

The working principle of hyperspectral image segmentation using EEOC is depicted in Figure 1.1. To begin with, the hyperspectral scene in the form of mat file is read; the feature matrix could be constructed based on mean or median absolute deviation (MAD), standard deviation (STD), variance (VAR). In addition, apply one of the clustering processes, namely K-means or Fuzzy C-means. Since these clustering method works depend on the number of cluster, Davies Bouldin (DB) index is used to determine the number of clusters. Furthermore, the dimensional reduction process could be carried out by this clustering method by picking out one band from each cluster, such as 204 bands of hyperspectral image are reduced to below twenty bands, i.e., one band is pick out based on maximum variance from each cluster. In particle swarm clustering algorithm, the input is reduced set of bands from K-means or FCM clustering algorithm. PSC (EEOC) is used for segmentation on the reduced bands.

Figure 1.1 Flowchart for hyperspectral image segmentation using EEOC.

1.3.2 Band Reduction Using K-Means Algorithm

K-means is an iterative technique that is used to partition the hyperspectral feature matrix into K clusters. It is one of the simplest unsupervised learning algorithms to solve the well-known clustering problem. The aim of this method is to minimize the sum of squared distances between all points and the cluster center.

The K-means algorithm is reduced the hyperspectral feature matrix is shown in Figure 1.2. Get the hyperspectral feature matrix and the number of clusters (determined from DB Index). Choose the K initial cluster centers, after which the new cluster is computed, 204 bands of the hyperspectral matrix are reduced to 17 bands, i.e., one band is selected according to the maximum variance from each cluster. Finally, the reduced set of bands is obtained.

Figure 1.2 Band reduction using K-means algorithm.

1.3.3 Band Reduction Using Fuzzy C-Means

Fuzzy C-means is an iterative technique that is used to partition the hyperspectral feature matrix into membership levels. Data elements can belong to more than one cluster in this method and associated with each element in a set of membership levels. It is a process of assigning these membership levels and then using them to assign data elements to one or more clusters.

The Fuzzy C-means algorithm is reduced the hyperspectral feature matrix is as abovementioned Figure 1.3. Get the hyperspectral feature matrix and the number of clusters that determine the DB Index. The membership matrix is initialized after which fuzzy cluster is calculated. Then, membership matrix is updated. 204 bands of the hyperspectral matrix are reduced to 17 bands i.e. one band is selected according to the maximum variance from each cluster. Finally, the reduced set of bands is obtained. The above flowchart shows the flow of band reduction using Fuzzy C-means clustering algorithm.

By using this above said clustering method, the hyperspectral bands are reduced into below 20 bands, depending on the scene. The DB index helps to determine the number of clusters, based on this number of cluster the bands are reduced, i.e., applying this clustering method into the hyperspectral feature matrix. These matrices are clusters that depend on the number of clusters. Then, pick out one band from each cluster, which has a maximum variance, i.e., number of clusters is 18, that implies the maximum variance band is pick out from each cluster and thus this hyperspectral data are reduced into 18 bands. This reduced band is allowed to segmentation, which is carried out by a light weight clustering algorithm. This algorithm is entitled as enhanced estimation of centroid (EEOC), which is the modification of the particle swarm clustering method performed. The modification is that the particle has updated their positions and computed the distance matrix only once per iteration.

Figure 1.3 Band reduction using Fuzzy C-means.

1.4 Experimental Results

The performance of EEOC algorithm is analyzed and experimented. Its performance is noticed to be satisfactory in term of reduction in the time complexity and the efficiency of each position update.

1.4.1 DB Index Graph

Figure 1.4(a) depicts DB Index Graph. In x-axis, number of clusters (i.e., the value is incremented by 1 pixel) and in y-axis, DB value (i.e., the value is increment by 0.05 pixels) is determined.

For Salinas_A scene, the DB-value is least at 17th cluster. So, the DB Index of this scene is considering as 17. For Indian Pines the DB value is least at 11th cluster. So, the DB Index is considered as 11 for this scene. For Salinas Valley Scene, the DB value is least at 10th cluster. So, the DB Index is considered as 10 for this scene. For Pavia University, the DB value is least at 20th cluster. So, the DB Index is considered as 20 for this scene. For Pavia Centre Scene, the DB value is least at 19th cluster. So, the DB Index is considered as 19 for this scene.

Figure 1.4 DB Index Graph. (a) Salina_A scene (b) Indian Pines (c) Salinas Valley (d) Pavia University and (e) Pavia Center scene.

1.4.2 K-Means–Based PSC (EEOC)

Figure 1.5(a) shows the segmentation result for the hyperspectral scene, namely Salinas_A, is processed by K-means based PSC (EEOC). In the interband clustering, K-means is utilized to reduce the band, whereas PSC (EEOC) is performed in intracluster. For this reason this approach is named as K-means based PSC (EEOC). Figures 1.6(a), 1.7(a), 1.8(a), 1.9(a) portrays the segmentation results for the hyperspectral scenes, namely Indian Pines, Salinas Valley, Pavia University, and Pavia Center, respectively. The pixels, which are clustered based on K-means working with PSC (EEOC), are displayed in Table 1.1.

Figure 1.5 Results for Salinas_A, (a) K-means + PSC (EEOC) (b) FCM + PSC (EEOC).

Table 1.1 Pixels clustered based on PSC (EEOC) with K-means and FCM for Salinas_A.

SALINAS _A 83 × 86 × 204

Clusters

K–Means + PSC (EEOC)

FCM + PSC (EEOC)

Pixels

Pixels

1

1284

1371

2

1492

2788

3

1081

507

4

611

1

5

507

1431

6

2163

1040

1.4.3 Fuzzy C-Means–Based PSC (EEOC)

Figures 1.5(b), 1.6(b), 1.7(b), 1.8(b), and 1.9(b) portray the segmentation results for FCM based on PSC (EEOC), which is examined on this HSI. The pixels are clustered based on FCM +PSC (EEOC) is listed in Table 1.1.

The following figures (Figures 1.5 to 1.9), left side images (a images), depict the results for K-means working with PSC (EEOC) of the various hyperspectral scenes. The right side listed images (b images) portray the results for FCM working with PSC (EEOC).

Figure 1.6 Results for Indian Pines (a) K-means + PSC (EEOC) (b) FCM +PSC (EEOC).

Figure 1.7 Results for Salinas Valley. (a) K-means + PSC (EEOC) (b) FCM + PSC (EEOC).

Figure 1.8 Results for Pavia University. (a) K-means + PSC (EEOC) (b) FCM + PSC (EEOC).

Figure 1.9 Results for Pavia Centre. (a) K-means + PSC (EEOC) (b) FCM + PSC (EEOC).

1.5 Analysis of Results

The following tables illustrate the resultant of the number of pixels, which are clustered. The number of the cluster is defined based on the ground truth, as well as important features of the abovementioned HSI.

Table 1.1 depicts the result of performance based on pixel wise cluster for the Salinas_A scene. As per the reference of the ground truth image, Salinas_A scene has segmented with six clusters. The size of this scene is 83 × 86 × 204. The 204 bands are reduced into 17 bands as per the Davies– Bouldin Index and finally this 83 x 86 pixels are clustered into six, such as 7138 pixels are grouped into these clusters. From Table 1.1, K-means working with PSC (EEOC) clustered the pixels equally, whereas 4th cluster has one pixel for FCM.

Table 1.2 portrays the result of performance based on pixel wise cluster for the Indian Pines scene. As per the reference of the ground truth image, the Indian Pines scene has segmented with seven clusters. The size of this scene is 145 × 145 × 200. The 200 bands are reduced into 11 bands as per the Davies-Bouldin Index and finally this 145 × 145 pixels are clustered into seven, such as 21,025 pixels are grouped into these clusters.

Table 1.3 portrays the result of performance based on pixel wise cluster for the Salinas Valley scene. As per the reference of the ground truth image, this scene has segmented with seven clusters. The size of this scene is 512 × 217 × 204. The 204 bands are reduced into 10 bands as per the Davies-Bouldin Index and finally this 512 × 217 pixels are clustered into seven, and 111,104 pixels are grouped into these clusters.

Table 1.2 Pixels clustered based on PSC (EEOC) with K-means and FCM for Indian Pines.

Indian Pines 145 × 145 × 200

Clusters

K-Means + PSC (EEOC)

FCM + PSC (EEOC)

Pixels

Pixels

1

2035

3354

2

2953

3522

3

1393

3700

4

3177

1497

5

2662

2193

6

3597

3603

7

5208

3156

Table 1.3 Pixels clustered based on PSC (EEOC) with K-means and FCM for Salinas Valley.

Salinas Valley 512 × 217 × 204

Clusters

K–Means + PSC (EEOC)

FCM + PSC (EEOC)

Pixels

Pixels

1

9006

22894

2

29424

21034

3

6695

5021

4

35776

11939

5

23797

18292

6

6328

7784

7

78

24140

Table 1.4 portrays the result of performance based on pixel wise cluster for the Pavia University scene. As per the reference of the ground truth image, this scene has segmented with nine clusters. The size of this scene is 610 × 340 × 103. The 103 bands are reduced into 20 bands as per the DB-Index, and finally these 610 × 340 pixels are clustered into nine, such as 207,400 pixels are grouped into these clusters.

Table 1.5 portrays the results of performance based on pixel wise cluster for the Pavia Centre scene. As per the reference of the ground truth image, Pavia Centre scene has segmented with nine clusters. The size of this scene is 1096 × 715 × 102. These 102 bands are reduced into 17 clusters, as per the Davies-Bouldin Index and finally this 1096 × 715 pixels are clustered into six, such as 783640 pixels are grouped into these clusters.

Table 1.6 depicted the time taken for execution of these methods. These approaches are applying on the above said hyperspectral scenes. From the table, K-means + PSC method takes minimum time to execute for all HSI, whereas FCM + PSC method takes more time for processed.

Table 1.4 Pixels clustered based on PSC (EEOC) with K-means and FCM for Pavia University.

Pavia University 610 × 340 × 103

Clusters

K-Means + PSC (EEOC)

FCM + PSC (EEOC)

Pixels

Pixels

1

2176

104374

2

656

4899

3

637

162

4

116561

23482

5

13031

1013

6

72789

53200

7

128

2150

8

610

2769

9

812

15351

Table 1.5 Pixels clustered based on PSC (EEOC) with K-means and FCM for Pavia Centre.

Pavia Center 1096 × 715 × 102

Clusters

K –Means + PSC (EEOC)

FCM + PSC (EEOC)

Pixels

Pixels

1

454182

38,965

2

22

156,770

3

52534

88,386

4

26

63,542

5

130076

194,464

6

9294

13,021

7

132714

115,264

8

20

113,205

9

4772

23

Table 1.6 Elapsed time in seconds for PSC (EEOC) with K-means and FCM.

Input

Size

K-means + PSC (EEOC)

FCM + PSC (EEOC)

SALINAS_A

83 × 86 × 204

85.4219s

102.5313s

INDIAN PINES

145 × 145 × 200

95.6563s

104.6719s

SALINAS VALLEY

512 × 217 × 204

277.8438s

283.7031s

PAVIA UNIVERSITY

610 × 340 × 103

9.98e+02s

10.48e+2s

PAVIA CENTER

1096 × 715 × 102

19.68e+2s

18.11e+2s

Since, in these works, particle swarm clustering method is used, fitness value is one of the parameters to measure the accuracy. The least fitness value indicates the optimum results.

Table 1.7 depicts that the fitness value of EEOC worked with K-means/FCM. PSC (EEOC) worked with FCM produce least fitness value other than the Salinas_A scene. This scene majorly contains the straight line. For this reason, EEOC worked with K-means produce optimum results. In addition, here, it is to be noted that the fitness value for K-means + PSC (EEOC) is a nearby triple of FCM + PSC (EEOC) for Pavia University. For Pavia Centre, the fitness value of FCM + PSC (EEOC) is more than double of K-means + PSC (EEOC). As a whole, EEOC is producing optimum results when worked with the abovementioned clustering methods.

Table 1.7 Fitness value for PSC (EEOC) with K-means and FCM.

Input

K-means + PSC (EEOC)

FCM + PSC (EEOC)

SALINAS_A

15.99e+5

17.96e+05

INDIAN PINES

96.17e+5

92.11e+05

SALINAS VALLEY

52.38e+6

40.82+06

PAVIA UNIVERSITY

174.56e+6

73.65e+06

PAVIA CENTER

34.16e+07

88.72e+07

1.6 Conclusions

An integrated image segmentation process based on interband clustering and intraband clustering is proposed. EEOC is used for segmentation process on the reduced bands. The hyperspectral bands are clustered, and a band which has a maximum variance from each cluster is chosen. K-means and FCM are used for interband clustering part, i.e., reduced the band. The Davies-Bouldin Index is used to determine the number of clusters. Finally, reduced set of bands are obtained. PSC (EEOC) takes the reduced set of bands as input and produces the segmentation result.