Machine Intelligence, Big Data Analytics, and IoT in Image Processing E-Book
173,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Wissenschaft und neue Technologien
- Sprache: Englisch
MACHINE INTELLIGENCE, BIG DATA ANALYTICS, AND IoT IN IMAGE PROCESSING
Discusses both theoretical and practical aspects of how to harness advanced technologies to develop practical applications such as drone-based surveillance, smart transportation, healthcare, farming solutions, and robotics used in automation.
The concepts of machine intelligence, big data analytics, and the Internet of Things (IoT) continue to improve our lives through various cutting-edge applications such as disease detection in real-time, crop yield prediction, smart parking, and so forth. The transformative effects of these technologies are life-changing because they play an important role in demystifying smart healthcare, plant pathology, and smart city/village planning, design and development. This book presents a cross-disciplinary perspective on the practical applications of machine intelligence, big data analytics, and IoT by compiling cutting-edge research and insights from researchers, academicians, and practitioners worldwide. It identifies and discusses various advanced technologies, such as artificial intelligence, machine learning, IoT, image processing, network security, cloud computing, and sensors, to provide effective solutions to the lifestyle challenges faced by humankind.
Machine Intelligence, Big Data Analytics, and IoT in Image Processing is a significant addition to the body of knowledge on practical applications emerging from machine intelligence, big data analytics, and IoT. The chapters deal with specific areas of applications of these technologies. This deliberate choice of covering a diversity of fields was to emphasize the applications of these technologies in almost every contemporary aspect of real life to assist working in different sectors by understanding and exploiting the strategic opportunities offered by these technologies.
Audience
The book will be of interest to a range of researchers and scientists in artificial intelligence who work on practical applications using machine learning, big data analytics, natural language processing, pattern recognition, and IoT by analyzing images. Software developers, industry specialists, and policymakers in medicine, agriculture, smart cities development, transportation, etc. will find this book exceedingly useful.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 771
Veröffentlichungsjahr: 2023
Ähnliche
Table of Contents
Cover
Series Page
Title Page
Copyright Page
Preface
Part I: DEMYSTIFYING SMART HEALTHCARE
1 Deep Learning Techniques Using Transfer Learning for Classification of Alzheimer’s Disease
1.1. Introduction
1.2. Transfer Learning Techniques
1.3. AD Classification Using Conventional Training Methods
1.4. AD Classification Using Transfer Learning
1.5. Conclusion
References
2 Medical Image Analysis of Lung Cancer CT Scans Using Deep Learning with Swarm Optimization Techniques
2.1. Introduction
2.2. The Major Contributions of the Proposed Model
2.3. Related Works
2.4. Problem Statement
2.5. Proposed Model
2.6. Dataset Description
2.7. Results and Discussions
2.8. Conclusion
References
3 Liver Cancer Classification With Using Gray-Level Co-Occurrence Matrix Using Deep Learning Techniques
3.1. Introduction
3.2. Related Works
3.3. Proposed Methodology
3.4. Conclusion
References
4 Transforming the Technologies for Resilient and Digital Future During COVID-19 Pandemic
4.1. Introduction
4.2. Digital Technologies Used
4.3. Challenges in Transforming Digital Technology
4.4. Implications for Research
4.5. Conclusion
References
Part II: PLANT PATHOLOGY
5 Plant Pathology Detection Using Deep Learning
5.1. Introduction
5.2. Plant Leaf Disease
5.3. Background Knowledge
5.4. Architecture of ResNet 512 V2
5.5. Methodology
5.6. Result Analysis
5.7. Conclusion
References
6 Smart Irrigation and Cultivation Recommendation System for Precision Agriculture Driven by IoT
6.1. Introduction
6.2. Related Works
6.3. Challenges of IoT in Smart Irrigation
6.4. Farmers’ Challenges in the Current Situation
6.5. Data Collection in Precision Agriculture
6.6. Conclusion
References
7 Machine Learning-Based Hybrid Model for Wheat Yield Prediction
7.1. Introduction
7.2. Related Work
7.3. Materials and Methods
7.4. Experimental Result and Analysis
7.5. Conclusion
Acknowledgment
References
8 A Status Quo of Machine Learning Algorithms in Smart Agricultural Systems Employing IoT-Based WSN: Trends, Challenges and Futuristic Competences
8.1. Introduction
8.2. Types of Wireless Sensor for Smart Agriculture
8.3. Application of Machine Learning Algorithms for Smart Decision Making in Smart Agriculture
8.4. ML and WSN-Based Techniques for Smart Agriculture
8.5. Future Scope in Smart Agriculture
8.6. Conclusion
References
Part III: SMART CITY AND VILLAGES
9 Impact of Data Pre-Processing in Information Retrieval for Data Analytics
9.1. Introduction
9.2. Related Work
9.3. Experimental Setup and Methodology
9.4. Experimental Result and Discussion
9.5. Conclusion and Future Work
References
10 Cloud Computing Security, Risk, and Challenges: A Detailed Analysis of Preventive Measures and Applications
10.1. Introduction
10.2. Background
10.3. Literature Review
10.4. Cloud Computing Challenges and Its Solution
10.5. Cloud Computing Security Issues and Its Preventive Measures
10.6. Cloud Data Protection and Security Using Steganography
10.7. Related Study
10.8. Conclusion
References
11 Internet of Drone Things: A New Age Invention
11.1. Introduction
11.2. Unmanned Aerial Vehicles
11.3. Application Areas
11.4. IoDT Attacks
11.5. Fusion of IoDT With Other Technologies
11.6. Recent Advancements in IoDT
11.7. Conclusion
References
12 Computer Vision-Oriented Gesture Recognition System for Real-Time ISL Prediction
12.1. Introduction
12.2. Literature Review
12.3. System Architecture
12.4. Methodology
12.5. Implementation and Results
12.6. Conclusion and Future Scope
References
13 Recent Advances in Intelligent Transportation Systems in India: Analysis, Applications, Challenges, and Future Work
13.1. Introduction
13.2. A Primer on ITS
13.3. The ITS Stages
13.4. Functions of ITS
13.5. ITS Advantages
13.6. ITS Applications
13.7. ITS Across the World
13.8. India’s Status of ITS
13.9. Suggestions for Improving India’s ITS Position
13.10. Conclusion
References
14 Evolutionary Approaches in Navigation Systems for Road Transportation System
14.1. Introduction
14.2. Related Studies
14.3. Navigation Based on Evolutionary Algorithm
14.4. Meta-Heuristic Algorithms for Navigation
14.5. Conclusion
References
15 IoT-Based Smart Parking System for Indian Smart Cities
15.1. Introduction
15.2. Indian Smart Cities Mission
15.3. Vehicle Parking and Its Requirements in a Smart City Configuration
15.4. Technologies Incorporated in a Vehicle Parking System in Smart Cities
15.5. Sensors for Vehicle Parking System
15.6. IoT-Based Vehicle Parking System for Indian Smart Cities
15.7. Advantages of IoT-Based Vehicle Parking System
15.8. Conclusion
References
16 Security of Smart Home Solution Based on Secure Piggybacked Key Exchange Mechanism
16.1. Introduction
16.2. IoT Challenges
16.3. IoT Vulnerabilities
16.4. Layer-Wise Threats in IoT Architecture
16.5. Attack Prevention Techniques
16.6. Conclusion
References
17 Machine Learning Models in Prediction of Strength Parameters of FRP-Wrapped RC Beams
17.1. Introduction
17.2. Strengthening of RC Beams With FRP Systems
17.3. Machine Learning Models
17.4. Conclusion
References
18 Prediction of Indoor Air Quality Using Artificial Intelligence
18.1. Introduction
18.2. Indoor Air Quality Parameters
18.3. AI in Indoor Air Quality Prediction
18.4. Conclusion
References
Index
End User License Agreement
List of Tables
Chapter 1
Table 1.1 AD classification via deep learning models.
Table 1.2 AD classification via transfer learning techniques.
Chapter 2
Table 2.1 Confusion matrix.
Table 2.2 Network model of the 2D CNN AlexNet.
Table 2.3 Network model of the 3D CNN AlexNet.
Table 2.4 Proposed CNN model.
Table 2.5 Performance metrics of the 3D-CNN classifier.
Table 2.6 Results comparisons with other classifiers.
Chapter 6
Table 6.1 Environmental considerations impacting different stages of product...
Chapter 7
Table 7.1 Input parameters in the wheat crop dataset.
Table 7.2 Few instances of wheat crop dataset.
Table 7.3 ReliefF algorithm determines the rank and weight for each feature....
Table 7.4 Experimental results for traditional ML techniques for wheat yield...
Table 7.5 Experimental results for each hybrid model for wheat yield predict...
Chapter 8
Table 8.1 Various sensors employed in smart agriculture.
Table 8.2 Machine learning solutions for various WSN issues.
Table 8.3 Analysis of different ML and WSN techniques for smart agriculture....
Chapter 9
Table 9.1 Description of PID attributes.
Table 9.2 Detailed description of PID attributes.
Table 9.3 The most prominent attributes from Cleveland dataset are extracted...
Table 9.4 The most prominent attributes from FHS are extracted using data pr...
Table 9.5 Description of attributes in the database.
Table 9.6 Performance of DT, NB and ANN on four distinct datasets before and...
Chapter 10
Table 10.1 Taxonomy of challenges in cloud security.
Table 10.2 Explanation of major security issues.
Table 10.3 Survey discussed in existing literature.
Chapter 12
Table 12.1 The comparative analysis of different studies in literature for S...
Chapter 13
Table 13.1 Major applications of ITS.
Table 13.2 Major cities using ITS.
Chapter 14
Table 14.1 Binary encoding.
Table 14.2 Permutation encoding.
Table 14.3 Value encoding.
Table 14.4 Tree encoding.
Table 14.5 Single point crossover (| is the crossover point).
Table 14.6 Mutation.
Table 14.7 Comparison between evolutionary algorithm and mathematical progra...
Chapter 15
Table 15.1 Minimum parking space for different types of buildings.
Table 15.2 Notable works in the literature regarding modernized vehicle park...
Chapter 17
Table 17.1 Mechanical characteristics of carbon fibers.
Table 17.2 ML research studies in FRP-concrete bond strength.
Table 17.3 Summary of ML research studies in shear strength prediction.
Table 17.4 Acronyms and symbols.
Chapter 18
Table 18.1 Sources of IAP and adverse effects of IAPs on health.
Table 18.2 Particulates classification based on their penetration level in t...
Table 18.3 Summary of use of AI in the prediction of IAQ.
List of Illustrations
Chapter 1
Figure 1.1 Transfer learning techniques.
Figure 1.2 TL process.
Chapter 2
Figure 2.1 Basic block diagram of CNN with PSO.
Figure 2.2 Proposed 3D CNN architecture.
Figure 2.3 Linear filter with bias.
Figure 2.4 Fully connected CNN.
Figure 2.5 Showing the CNN with filter and stride.
Figure 2.6 AUC-ROC curve of the proposed algorithm.
Figure 2.7 Training image of the Luna-16 dataset.
Chapter 3
Figure 3.1 Basic anatomy of human liver [1].
Figure 3.2 Basic flow of the pre-processing.
Figure 3.3 Proposed flow diagram for liver cancer classification.
Figure 3.4 Basic CT scan of the liver with enhanced multiphase.
Figure 3.5 CNN-based classified image.
Figure 3.6 Confusion matrix of the ROC curve.
Figure 3.7 Classified output after applying proposed algorithm.
Chapter 4
Figure 4.1 Digital technologies.
Figure 4.2 Challenges of digital technology.
Chapter 5
Figure 5.1 Classification of plant leaf diseases.
Figure 5.2 ResNet 152 V2 architecture.
Figure 5.3 Residual learning process.
Figure 5.4 Proposed method for detecting healthy leaf or unhealthy leaf.
Figure 5.5 Data normalization.
Figure 5.6 95% accuracy obtained with ResNet152V2.
Figure 5.7 Curve for model loss accuracy.
Figure 5.8 Curve for test accuracy.
Chapter 6
Figure 6.1 IoT analytics life cycle.
Figure 6.2 Ratio of rainfall versus humidity across various crops.
Figure 6.3 Challenges related to IoT in smart irrigation.
Figure 6.4 Precision agriculture cycle in IoT.
Figure 6.5 Proposed model and its working.
Figure 6.6 Data flow of smart irrigation.
Chapter 7
Figure 7.1 Methodology for the proposed work.
Figure 7.2 Weight value for each feature as predicted by ReliefF algorithm....
Figure 7.3 General architecture of ANN.
Figure 7.4 Comparative analysis of traditional KNN vs hybrid models.
Figure 7.5 Comparative analysis of traditional SVM vs hybrid models.
Figure 7.6 Comparative analysis of traditional LR vs hybrid models.
Figure 7.7 Comparative analysis of traditional ANN vs hybrid models.
Figure 7.8 Comparative analysis of traditional LDA vs hybrid models.
Figure 7.9 Comparative analysis of comparative analysis of traditional NB vs...
Chapter 8
Figure 8.1 Classification of WSN issues.
Chapter 9
Figure 9.1 Stages of data pre-processing.
Figure 9.2 Different tasks in data pre-processing [7].
Figure 9.3 Performance of DT, NB and ANN on PIMA dataset before and after pr...
Figure 9.4 Variations in results before and after the pre-processing of the ...
Figure 9.5 Performance of DT, NB and ANN on Cleveland dataset before and aft...
Figure 9.6 Variations in results before and after the pre-processing of the ...
Figure 9.7 Performance of DT, NB and ANN on FHS dataset before and after pre...
Figure 9.8 Variations in results before and after the pre-processing of the ...
Figure 9.9 Performance of DT, NB and ANN on diabetic dataset before and afte...
Figure 9.10 Variations in results before and after the pre-processing of the...
Chapter 10
Figure 10.1 Beginning era of cloud.
Figure 10.2 Types of cloud services and architecture.
Figure 10.3 Cloud computing services.
Figure 10.4 Application areas of cloud computing.
Figure 10.5 Cloud computing architecture as per cloud-server communication....
Figure 10.6 Cloud security issues.
Figure 10.7 Steganography process.
Figure 10.8 Steps of steganography data encryption.
Figure 10.9 Color bit representation.
Chapter 11
Figure 11.1 Working of UAV.
Figure 11.2 IoDT architecture.
Figure 11.3 IoDT applications.
Figure 11.4 IoDT attacks.
Chapter 12
Figure 12.1 (a) Characters Dataset [A-R] before pre-processing, (b) Characte...
Figure 12.2 System architecture for the proposed ISL Model.
Figure 12.3 CNN model of the proposed system.
Figure 12.4 Output character display of numeric values within the range of “...
Figure 12.5 Output character display of alphabets within the range of “A–J.”...
Figure 12.6 Loss-accuracy curves achieved for the ISL model.
Figure 12.7 Heat map for confusion matrix for the trained model.
Chapter 13
Figure 13.1 Stages of ITS.
Figure 13.2 Functions of ITS.
Figure 13.3 Advantages of ITS.
Chapter 14
Figure 14.1 Block diagram of global navigation satellite system.
Figure 14.2 Trilateration of GPS.
Figure 14.3 Segment of GPS.
Figure 14.4 Taxonomy of evolutionary algorithms.
Figure 14.5 Flow of differential evolution and genetic algorithm.
Figure 14.6 Tournament selection.
Figure 14.7 Roulette wheel selection.
Figure 14.8 Cast of characters for genetic algorithm and differential evolut...
Figure 14.9 Illustration of TSP.
Chapter 15
Figure 15.1 Smart city components.
Figure 15.2 Parking planning cycle.
Figure 15.3 Infrared sensor in parking systems.
Figure 15.4 Ultrasonic sensors in the vehicle parking system.
Figure 15.5 CCTV implementation in smart parking management.
Figure 15.6 Special parking slots for inner-city traffic.
Figure 15.7 Vehicle management system.
Figure 15.8 Vehicle parking system using IoT devices.
Chapter 16
Figure 16.1 Generic IoT architecture.
Figure 16.2 Layer-wise IoT device architecture.
Figure 16.3 Key exchange mechanism for IoT devices.
Chapter 17
Figure 17.1 Types of FRPs (a) carbon, (b) glass (c) aramid (d) basalt.
Figure 17.2 Structure of FRP system.
Figure 17.3 Types of strengthening in RC beams with FRP.
Figure 17.4 FRP-bond strength test setup (a) single-lap shear test [19] (b) ...
Figure 17.5 Flexural reinforcing of beams with EB FRP composite (a) bottom f...
Figure 17.6 FRP shear strengthening configurations [23].
Figure 17.7 AI in the prediction of FRP strengthened RC beams.
Figure 17.8 R
2
value of different ML models (FRP-concrete bond strength).
Figure 17.9 R
2
value of different ML models (shear strength of FRP strengthe...
Chapter 18
Figure 18.1 IAQ parameters.
Figure 18.2 Indoor air pollutants and their particle size range.
Figure 18.3 AI, ML, and DL.
Figure 18.4 ML methods.
Guide
Cover Page
Series Page
Title Page
Copyright Page
Preface
Table of Contents
Begin Reading
Index
WILEY END USER LICENSE AGREEMENT
Pages
ii
iii
iv
xv
xvi
xvii
xviii
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
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
146
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
172
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
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
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
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
Scrivener Publishing100 Cummings Center, Suite 541JBeverly, MA 01915-6106
Advances in Intelligent and Scientific Computing
Series Editors: Dr. Sujata Dash, Dr. Subhendu Kumar Pani and Dr. Milan Tub
The series provides in-depth coverage of innovations in artificial life, computational intelligence, evolutionary computing, machine learning and applications. It is the intention for the volumes in the series to be practically relevant, so that the results will be useful for managers in leadership roles. Therefore, both theoretical and managerial implications of the research will be considered.
Submission of book proposals toDr. Subhendu Kumar Pani at [email protected] [email protected]
Publishers at ScrivenerMartin Scrivener ([email protected])Phillip Carmical ([email protected])
Machine Intelligence, Big Data Analytics, and IoT in Image Processing
Practical Applications
Edited by
Ashok Kumar
Chitkara University Institute of Engineering & Technology, Chitkara University, Punjab, India
Megha Bhushan
School of Computing, DIT University, Dehradun, Uttarakhand, India
José A. Galindo
Department of Computer Languages and Systems, University of Seville, Spain
Lalit Garg
Computer Information Systems, University of Malta, Malta
and
Yu-Chen Hu
Dept. of Computer Science and Information Management, Providence University, Tai Chung, Taiwan
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 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 orga-nization, 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 978-1-119-86504-9
Cover image: Pixabay.ComCover design by Russell Richardson
Preface
The concepts of machine intelligence, big data analytics and the Internet of Things (IoT) continue to improve our lives through various cutting-edge applications such as disease detection in real time, crop yield prediction, smart parking and so forth. The transformative effects of these technologies are life-changing because they play an important role in demystifying smart healthcare, plant pathology, and smart city/village planning, design and development. This book presents a cross-disciplinary perspective on the practical applications of machine intelligence, big data analytics and IoT by compiling cutting-edge research and insights from researchers, academicians and practitioners worldwide. It identifies and discusses various advanced technologies, such as artificial intelligence, machine learning, IoT, image processing, network security, cloud computing and sensors, to provide effective solutions to the lifestyle challenges faced by humankind.
These practical innovative applications may include navigation systems for road transportation, IoT- and WSN-based smart agriculture, plant pathology through deep learning, cancer detection from medical images and smart home solutions. Moreover, cloud computing has made it possible to access these real-life applications remotely over the internet. The primary concern of this book is to equip those new to this field of application, as well as those with more advanced knowledge related to practical application development, exploit the inherent features of machine intelligence, big data analytics and IoT. For instance, how to harness these advanced technologies to develop practical applications such as drone-based surveillance, smart transportation, healthcare, smart farming solutions, and robotics for automation.
This book is a significant addition to the body of knowledge on practical applications emerging from machine intelligence, big data analytics and IoT. The chapters deal with specific areas of applications of these technologies. This deliberate choice of covering a diversity of fields was to empha-size the applications of these technologies in almost every contemporary aspect of real life to assist working in different sectors by understanding and exploiting the strategic opportunities offered by these technologies. A summary of the main ideas of the work presented in each of the chapters follows:
Chapter 1
is based on the models used to diagnose Alzheimer’s disease (AD). These models utilize CaffeNet, GoogLeNet, VGGNet-16, VGGNet-19, DenseNet with varying depths, Inception-V4, AlexNet, ResNet-18, ResNet-152, or even ensemble transfer-learning, that are pre-trained on generalized images for AD classification to achieve better performance as compared to training a model from scratch.
Chapter 2
describes how to detect cancerous lung nodules from a lung CT scan image given as input and how to classify the lung cancer along with its severity. A novel deep learning method is used to detect the location of cancerous lung nodules.
Chapter 3
outlines a classifier used to divide the liver and CT images into normal and abnormal categories based on the main features in terms of shape, texture, and feature statistics. It includes four stages: preprocessing, fuzzy clustering, feature extraction and classification. Furthermore, the grey-level co-occurrence matrix (GLCM) method is used to extract the features.
Chapter 4
provides some of the major emerging digital technologies which have transformed the lives of individuals by making their future dependent upon the resilience of these technologies. It also highlights some of the major challenges related to these technologies with their suitable implications.
Chapter 5
describes a model based on ResNet architecture in deep learning to help farmers detect plant leaf diseases at an early stage in order to take precautionary measures against them.
Chapter 6
discusses an IoT-based smart irrigation system to assist farmers in precision agriculture for increasing crop yield. It uses multiple sensor metrics to help anticipate conditions for irrigation planning by predicting soil moisture, temperature, and humidity.
Chapter 7
presents a hybrid model for wheat crop yield prediction using machine learning (ML) approaches, namely k-nearest neighbors (KNN), naïve Bayes, artificial neural network, logistic regression, support vector machine and linear discriminant analysis. The model works in two stages: the first stage uses a feature selection strategy to find the best features for wheat crops, and the second stage uses ML to estimate crop yield based on these best features.
Chapter 8
discusses wireless sensor network (WSN)-based techniques used for smart agriculture and applications of ML for smart decision-making.
Chapter 9
provides an insight into the applications of data preprocessing techniques and their effects on information retrieval. It covers the major issues that need to be dealt with before beginning any data analysis process.
Chapter 10
focuses on the security for the latest paradigm shift in cloud and distributed computing. It delineates various risk parameters in the cloud environment and provides some novel methods to be adopted for cloud data security.
Chapter 11
talks about the internet of drone things (IoDT), its applications in the modern world, research opportunities, and current challenges to be dealt with. Furthermore, it discusses new age inventions, security issues, and attacks that frequently occur in the IoDT.
Chapter 12
presents an artificial intelligence-based gesture recognition system for the prediction of Indian sign language in real time. It covers different experiments using two-dimensional convolutional neural network-based classification to convert images into text.
Chapter 13
sets forth applications, challenges, and future developments in the field of intelligent transportation systems (ITS) in India. It explains ITS and evaluates their feasibility in India.
Chapter 14
provides a survey of evolutionary techniques used in navigation to create opportunities for analysts and researchers seeking to understand the broad pattern of different algorithms used in the navigation system.
Chapter 15
examines the IoT-based vehicle parking system in Indian cities. Additionally, it discusses vehicle parking and its basic requirements, various technologies incorporated in modern parking systems, different sensors utilized in parking facilities, and the advantages of IoT-based vehicle parking systems in detail.
Chapter 16
discusses a secure data transmission and key exchange for ensuring the confidentiality of data. Also, a lightweight authentication mechanism for ensuring the integrity and confidentiality of data shared over an unsecured network is presented.
Chapter 17
delineates machine learning models in the prediction of strength parameters of fiber-reinforced polymer (FRP) wrapped reinforced concrete (RC) beams. It provides a summary of machine learning models in the estimation of bond strength between FRP and concrete surface, and shear and flexural strength of FRP wrapped RC beams.
Chapter 18
describes existing AI-based studies for forecasting the indoor air quality of buildings and the future of AI-based indoor air quality forecasting. It provides an overview of the important role of machine learning models in the prediction of indoor pollutant concentrations to develop warning systems which help to affect the occupant’s health positively.
This book was edited by a team of academicians and experts. It is our hope that readers will draw several benefits from both the theoretical and practical aspects covered in the book to enhance their own practice or research.
The Editors
Dr. Ashok Kumar
Phagwara, India
Dr. Megha Bhushan
Dehradun, India
Dr. José Galindo
Seville, Spain
Dr. Lalit Garg
Valetta, Malta
Dr. Yu-Chen Hu
Tai Chung, Taiwan
January 2023
Part IDEMYSTIFYING SMART HEALTHCARE
1Deep Learning Techniques Using Transfer Learning for Classification of Alzheimer’s Disease
Monika Sethi1, Sachin Ahuja2* and Puneet Bawa1
1 Chitkara University Institute of Engineering & Technology, Chitkara University, Punjab, India
2 ED-Engineering at Chandigarh University, Punjab, India
Abstract
Alzheimer’s disease (AD) is a severe disorder in which brain cells degenerate, increasing memory loss with treatment choices for AD symptoms varying based on the disease’s stage, and as the disease progresses, individuals at certain phases undergo specific healthcare. The majority of existing studies make predictions based on a single data modality either they utilize magnetic resonance imaging (MRI)/positron emission tomography (PET)/diffusion tensor imaging (DTI) or the combination of these modalities. However, a thorough understanding of AD staging assessment can be achieved by integrating these data modalities and performance could be further enhanced using a combination of two or more modalities. However, deep learning techniques trained the network from scratch, which has the following drawbacks: (a) demands an enormous quantity of labeled training dataset that could be a problem for the medical field where physicians annotate the data, further it could be very expensive, (b) requires a huge amount of computational resources. (c) These models also require tedious and careful adjustments of numerous hyper-parameters, which results to under or overfitting and, in turn, to degraded performance. (d) With a limited medical training data set, the cost function might get stuck in a local-minima problem. In this chapter, a study is done based on the models used for AD diagnosis. Many researchers fine-tuned their networks instead of scratch training and utilized CaffeNet, GoogleNet, VGGNet-16, VGGNet-19, DenseNet with varying depths, Inception-V4, AlexNet, ResNet-18, ResNet-152, or even ensemble transfer-learning models pretrained on generalized images for AD classification performed better.
Keywords: Alzheimer disease, transfer learning, deep learning, parameter optimization
1.1. Introduction
In the United States, AD is the most widespread neurodegenerative condition and the sixth major cause of fatalities. The global disease burden of AD is expected to exceed $2 trillion by 2030, requiring preventative care [1]. Despite the tremendous study and advancements in clinical practice, nearly half of AD patients are correctly identified for anatomy and progression of the disease based on medical indicators. The existence of neu-rofibrillary tangles and amyloid plaques in histology is the most definitive evidence for AD. Consequently, the presence of plaque is not associated with the beginning of AD, but rather with sensory and neuron damage. Dr. Alois Alzheimer (a psychiatrist and neuropsychologist) was the origin for the naming of this disease, who studied the brain of a 51-year-old woman who died of severe cognitive impairment in 1906 [2]. Dr. Alois investigated her brain and discovered clumps, which were actually the accumulation of proteins in and around the neurons, resulting in their loss. The key characteristics for identifying or confirming the existence of the illness are shrinkage of the hippocampus and cerebral cortex, as well as growth of the ventricles. The hippocampus is essential in learning and memory, in addition to acting as a connection between the central nervous system of the body’s organs. AD eventually destroys the portion of the brain that controls heart and respiratory activity, resulting in death [3].
Unfortunately, AD does not yet have a definitive cure [4]. Instead, the objective is to reduce the illness’s development, treat suffering, manage learning disabilities, and enhance the quality of life. Clinical trials, on the other hand, can significantly slow down the progression of psychiatric disorders if the diagnosis is made early. Whereas more psychological therapies and, eventually, prevention or even a cure are essential (long-term) goals, early diagnosis may result in improved treatment outcomes benefits for diseased. Except in a few cases where genetic abnormalities may be identified, the precise cause of AD is still obscure.
The assessment of empirical biomarkers is necessary for the early treatment of disease [5]. A number of noninvasive neuroimaging approaches, including computed tomography (CT) scans, both structural and functional MRI and PET, have been explored for the prediction of AD. To produce cross sectional pictures of the bones, blood arteries, and soft tissues within the body, computer processing is used to integrate a succession of X-ray images recorded from different angle defined on your body. Plain X-rays do not give as much detail as CT scan imaging. An MRI scan employs a powerful magnet and radio waves to see at structures beneath the brain, according to the National Institute of Health. MRI scans are used by healthcare physicians to examine a variety of diseases, from damaged ligaments to cancer. To see and evaluate changes in cellular metabolism, PET is a functional imaging method, which thus employs radioactive additives termed as radiotracers.
Radiologists and clinicians, who are medical experts, analyze medical imaging data [6]. As a result of the probable tiredness of human specialists while evaluating images manually, a computer-assisted approach has proven to be beneficial for researchers as well as physicians. However, machine learning (ML) approaches are helping to improve the issue. Medical image analysis tasks need the use of ML to discover or learn useful features that characterize the correlations or patterns present in data. Since relevant or task-related characteristics are often created by human specialists on the basis of their domain expertise, it might be difficult for nonexperts to use ML techniques for their own research in the traditional manner. A number of projects are now working on the problem of learning sparse representations from training samples or pre-set dictionaries. Since then, there are attempts to generate sparse representations based on predefined dictionaries, which might be learned from training dataset. As a result of the concept of parsimony, sparse representation is used in many scientific fields. A sparsely inducing penalization and feature learning technique has been shown to be effective in medical image analysis when it comes to determining feature representation and selection [7]. Though data with a shallow architecture are still found to have meaningful patterns or regularities, techniques such as sparse representation or dictionary learning are still limited in their ability to represent them. Feature engineering has been incorporated into a learning phase in deep learning (DL), though, overcoming this issue [8]. Instead of manually extracting features, DL takes simply a collection of data with little preparation, if required, and then learns the valuable interpretations in an automatic method. Due to this shift in responsibility for feature extraction and selection, even non-experts in ML may now use DL effectively for their own research work, especially in the medical field for imaging analysis [9].
However, DL is afflicted by data dependency, one of the most signifi-cant problems. Comparatively to standard ML approaches, DL relies on a significant quantity of training data in order to discover hidden patterns in data. There is an interesting relationship between the size of the model in terms of the numbers of layers and the volume of information required.
Transfer learning (TL) eliminates the dependency of a huge amount of data requirement, which inspires us to utilize this to combat the problem of inadequate training data. This concept is driven by the idea that people may strategically utilize past knowledge to solve new problems or accomplish desirable results. The fundamental reasoning underlying this idea in ML was presented during a Neural Information Processing Systems (NIPS-95) symposium on “Learning to Learn,” which emphasized the need of lifelong ML approaches that store and apply previously acquired information [10]. TL approaches have recently shown results in a variety of practical applications. In Verma et al. [11], researchers utilized TL methods to transfer text data across domains. For fixing natural language processing issues, structural correspondence learning was presented by an author in Nalavade et al. [12]. Researchers employed several Convolutional Neural Network (CNN)-based TL models to detect AD [13].
This chapter presents the results of several TL techniques employed by previous researchers to identify AD.
1.2. Transfer Learning Techniques
TL is an ML research subject that focuses on retaining information received while addressing the problem and adapting it to some other but similar issue. As an instance, knowledge acquired when learning to identify trucks may be used while aiming to classify other four-wheeler vehicles. In CNN, this may be implemented in one of two ways: either the weights of all CNN layers are coupled to some other CNN layer with classification Layer output, as well as just utilizing “off-the-shelf CNN features,” whereby CNN serves like a generalized feature extractor to be analyzed later.
Several domains of knowledge engineering, such as classifier, prediction, and segmentation, have already experienced significant results using ML and data mining techniques [14]. Many ML techniques, however, operate successfully with an assumption that training test data are collected from the same dimensional region and variance. Most statistical models must still be redesigned from beginning when the population varies, employing new received data for training. In several practical applications, recollecting the necessary training data and rebuilding the models is either too expensive or not feasible. It would be extremely beneficial if researchers could reduce the need for the time and efforts associated with acquiring training samples. Transferring information or learning across problem contexts would be advantageous in such scenarios.
Researchers face three (1H and 2W) primary research issues in TL: how, when, and what to transfer [15]. What knowledge may be transmitted between areas or tasks is studied under what to transfer? There are certain types of knowledge that are particular to certain domains or tasks, and there are other types of information that are common to several domains and may serve to increase performance in the target task. To answer the “how to transfer” question, learning algorithms must be built after determining which types of knowledge may be transmitted. When to transfer skills is a question that explores when skill knowledge transfer will be performed. In the same way, researchers are curious about when learning should not be revealed. Different TL strategies are used to decide on 1H and 2W as illustrated in Figure 1.1.
In the case of inductive TL, whether the source and target domains are dissimilar or similar, the targeted task is independent of the input space. However, the source and target domains are distinct in a transductive TL situation. In case of unsupervised transfers, the destination task may be distinct from the source [16]. TL allows investigators to apply knowledge gained from previously completed work to related and newer situations. When researchers have a large amount of Learning Task 1 data from Source, they may use TL methodologies to acquire and generalize such Gathered Knowledge (properties, weights) with the goal of Learning Task 2, which has far less data as observed in Figure 1.2.
Figure 1.1 Transfer learning techniques.
The shorter training timeframes, improved neural-network efficiency (in most circumstances), as well as the lack of a large quantity of Source Dataset are only a several of the advantages of TL [17]. Similarly, TL in machine learning involves the application of previously learnt models to a new problem. When developing a computational model from the beginning, a considerable quantity data are usually necessary, but access to that knowledge is not always possible—this is when TL gets in handy. Further, instead of training the model from start to finish, remove the last fully connected output layer and utilize the pretrained computational model as a feature extractor [18, 19]. Similarly, we may utilize a new dataset to solve a different problem using similar strategy. One may develop a convenient and rapid linear model to change the output based on the new dataset since the pretrained advanced artificial neural network is used as a characteristic for the new task. Whenever the target task data is limited, the feature extraction technique is the best option. Fine-tuning is accomplished by unfreezing the underlying model (or a portion of it) and retraining the entire model on the full dataset at a low learning speed [20]. Effectiveness of the model in terms of accuracy performance somewhat on the given new dataset will be improved because to the modest learning rate, which also helps in minimizing over fitting in limited data scenarios. The learning rate has to be low since the model is large and the dataset is small. This is an overfitting equation that explains why and how the learning rate is very low. In such scenario, one would need to recompile the model that they have made these adjustments for them to take effect [21]. This is because the compile function locks the behaviour every time it is invoked. This means that if you want to change the model’s behavior, you will have to recompile it. The model will be retrained and monitored by call back to ensure that it does not over fit.
Figure 1.2 TL process.
1.3. AD Classification Using Conventional Training Methods
As a result of technological systems, e-health has gained remarkable progress. In addition to clinical predictive analytics, tele-health and patient monitoring tools are also examples of this technology. It is crucial to be able to predict and diagnose disease before it occurs. Psychiatric evaluation and smart watch data can be used by doctors to provide a diagnosis for the patient or forecast the possibility of future disease to aid patients postpone and avoid illnesses. Several diseases, including AD, are extremely challenging to diagnose in the initial phase because of their mild symptoms [22]. The usage of artificial intelligence systems centred on neural networks in patient care has steadily increased over the last decade [23]. These systems are now utilized in a variety of fields, such as medical assessment, cate-gorization, and forecasting domains. In an artificial neural network, the computation is performed in a distributed and parallel manner using many integer cores. Patterns are learnt through training samples and depending on the gained knowledge throughout training then applied to new unseen data. Table 1.1 shows the performance of several AD classification models utilized by the scientific community.
Table 1.1 AD classification via deep learning models.
Article reference
Objective (binary/ternary classification)
Dataset
Architecture
Classification (accuracy in %)
[
24
]
Binary
ADNI
Sparse Autoencoder with CNN
AD v/s HC (93.8)
AD v/s. MCI (86.3)
MCI v/s HC (83.3)
[
25
]
Binary
ADNI
Autoencoder stack with softmax Regression
AD v/s NC (88)
NC v/s MCI (77)
[
26
]
Binary
ADNI
Deep Boltzmann Machine
AD v/s NC (93.5)
MCI v/s NC (85)
cMCI v/s ncMCI (74.5)
[
27
]
Binary and Ternary
ADNI
2D-CNN
3D- CNN
2D-CNN
AD v/s HC (95.39)
HC v/s MCI (90)
AD v/s MCI (82)
AD v/s MCI vs. HC (86)
3D-CNN
AD v/s HC (95.39)
HC v/s MCI (92)
AD v/s MCI (87)
AD v/s MCI vs. HC (89)
[
28
]
Binary
-
CNN
AD vs. NC (86.7)
[
29
]
Binary
ADNI
Ensemble DBM
AD v/s NC (90)
MCI v/s AD (84)
MCI v/s NC (83)
[
30
]
Binary
ADNI
Cascaded 3D-CNN
AD v/s HC (92)
[
31
]
Binary
ADNI
CNN
AD v/s NC (Sensitivity-.69, Specificity-.98)
[
32
]
Binary
ADNI
ensemble CNN
AD v/s NC (89.60)
[
33
]
Binary
ADNI
3D CNN
AD v/s NC (80)
lMCI v/s eMCI (52)
[
34
]
Binary
ADNI
Multiple 3D CNN
AD v/s NC (92.26)
[
35
]
Binary
ADNI
CNN
AD v/s NC (79.9)
[
36
]
Binary
ADNI
2D CNN with RNN
AD v/s NC (95.3)
NC v/s MCI (83.9)
[
37
]
Multiclass Classification
ADNI
SAE and SVM
With SAE
AD v/s cMCI v/s ncMCI v/s NC (53)
With SVM
AD v/s cMCI v/s ncMCI v/s NC (47)
[
38
]
Binary
ADNI
CNN
AD v/s CN (90.1)
CN v/s pMCI (87.46)
pMCI v/s sMCI (76.9)
[
39
]
Binary
ADNI
3D CNN with LSTM
AD v/s NC (94.8)
pMCI v/s NC (86.3)
sMCI v/s NC (65.3)
[
40
]
Binary
ADNI
CNN
AD v/s NC (92.3)
AD vs. MCI (85.6)
MCI v/s NC (78.1)
[
41
]
Binary
ADNI
2D CNN, 3D CNN, 3D CNN SVM
NC v/s MCI (98.8)
NC v/s AD (99)
MCI v/s AD (89.4)
[
42
]
Binary
ADNI
3D CNN
AD v/s MCI (89.3)
MCI v/s NC (87.5)
[
43
]
Multiclass Classification
OASIS
CNN
NonD v/s VMild v/s Mild v/s Mod (71)
For binary classification, in 2014, the researchers employed SAEs with a regression layer soft max as activation function and reached an accuracy of 88% for AD v/s NC and 77% for MCI v/s NC [2]. With DBM architecture [26], researchers attained an accuracy of approximately 93% for AD vs. NC on the ADNI. There was a comparison between 2D CNN and 3D CNN for both binary as well as ternary AD classification, and 3D CNN model outperformed 2D CNN for both binary and ternary classification [3]. The accuracy of AD vs. CN for multiple 3D CNN was 92% in [8], while 2D CNN performed better than 3D CNN utilising RNN with an accuracy of 95.3% in [10]. For multiclass classification, only 71% accuracy was achieved [16] using the CNN architecture on OASIS dataset. Overall, 3D CNN performed better than 2D CNN, and the best accuracy was attained in the AD vs. NC classification, when compared to other binary and ternary classes.
1.4. AD Classification Using Transfer Learning
Unlike traditional ML, DL enables for automatic feature extraction from low to high level. In contrast, the network is trained from scratch using DL methods, which has certain disadvantages. In the medical area where clinicians annotate the data, this may be an issue and highly expensive. It also requires a significant amount of computing resources. In addition, several hyper-parameters in these models must be carefully adjusted, which could cause to under or over fitting, and degradation in performance. Due to the lack of data, the cost function may become stuck in ‘local minima’. TL may be used to fine-tune a deep network (such as CNN) rather than learning from scratch. A neural network model is initially trained on data from a source domain that is similar to the problem being targeted. It then uses the top few layers from the trained model to create a new one that uses the target dataset as its training data set. In order to perform TL, weight initialization and feature extraction are the two most important approaches. As a result, this approach is far quicker and produces better results than training a network from scratch. As a result of its huge computational capacity, it is mostly used for computer vision applications, such as emotional analysis and classification issues. Many research teams fine-tuned their networks instead of starting from scratch and used CaffeNet, GoogLeNet, VGGNet-11, VGGNet-16, DenseNet with varying depths (121-161-169), Inception-V4, AlexNets, ResNet-18, ResNet-152, or even ensemble TL models that are pretrained on general-ized images for the AD classification and improved performance. Table 1.2 illustrates the success of various designs via TL techniques utilized by researchers to classify AD.
Ternary AD classification using VGG-16 architecture was reported to be 92% accurate in [44]. On the other hand, VGG-based AD binary classification was shown to be more accurate by the authors [50, 55, 58, 61]. For ternary classification, GoogleNet outperformed ResNet-18 and ResNet-152 in [45] on ADNI dataset. There was a comparison of four different TL models (VGG, DenseNet, ResNet, and EfficientNet) in [54] using the same OASIS dataset and EfficientNet had the greatest accuracy of 96% out of four.
Table 1.2 AD classification via transfer learning techniques.
Article reference
Objective (binary/ternary/multi-class classification)
Dataset
Architecture
Classification (accuracy in %)
[
44
]
Ternary
ADNI
VGGNet
AD v/s MCI v/s HC (92)
[
45
]
Binary
ADNI
LeNet
AD v/s HC (98.84)
[
46
]
4-Way
ADNI
GoogleNet
ResNet-18
ResNet-152
AD v/s MCI v/s LMCI v/s HC
98.9
98.01
98.14
[
47
]
Multiclass Classification
OASIS
CNN (Hyperparameters of the Inception-V4 model)
NonD v/s VMild v/s Mild v/s
[
48
]
Ternary
ADNI
GoogleNet
CaffeNet
sMCI v/s cMCI v/s HC (83.23)
87.78
[
49
]
Multiclass Classification
OASIS
Inception-V4
ResNet
NonD v/s VMild v/s Mild v/s Mod (95)
[
50
]
Binary
ADNI
VGG
AD v/s NC (96.81)
MCI v/s NC (92.62)
[
51
]
Binary
ADNI
VGG-16
AD v/s CN vs. MCI (95.73)
[
52
]
Ternary
OASIS
Ensemble
ResNet50
DenseNet
AD v/s MCI vs. CN (95.23)
[
53
]
Binary
OASIS
Siamese CNN (ResNet-34)
AD v/s NC (98.72)
[
54
]
Binary
ADNI
AlexNet GoogLeNet ResNet
Highest Accuracy achieved in AlexNet then ResNet.
[
55
]
Binary
ADNI
VGG
AD v/s NC (98.7)
EMCI v/s LMCI (83.7)
[
56
]
Ternary
ADNI
AlexNet
AD v/s NC (97.2)
[
57
]
Binary
ADNI
ResNet-18
AD v/s NC (96.9)
[
58
]
Binary
ADNI
VGG-16
NC v/s AD (99.01)
MCI v/s AD (98.71)s
[
59
]
Multiclass Classification
OASIS
VGG
DenseNet
ResNet
EfficientNet
VGG
NonD v/s VMild v/s Mild v/s Mod (79)
DenseNet
NonD v/s VMild v/s Mild v/s Mod (92)
ResNet
NonD v/s VMild v/s Mild v/s Mod (93)
EfficientNet
NonD v/s VMild v/s Mild v/s Mod (96)
[
60
]
Multiclass Classification
OASIS
AlexNet
NonD v/s VMild v/s Mild v/s Mod (95)
[
61
]
Binary
ADNI
VGG-16
AD v/s CN (99.95)
As a whole, it is suggested to combine pretrained CNNs for initialization and retrain them utilising just fine-tuning of the CNNs layers for better performance for AD classification.
1.5. Conclusion
This chapter discussed binary and multiclass classification approaches using conventional and DL models using TL for AD classification. AD is a leading risk factor for death in developed nations. From a scientific perspective, computer-aided algorithms have yielded excellent findings, but practically, there is still no viable diagnostic technique usable. In the past few years, DL models have grown increasingly popular, especially when it comes to AD classification. A deep model was trained from scratch in the majority of the experiments, although this is sometimes impractical because the training procedure is time-consuming and a significant bit of training dataset is necessary to make it work efficiently. For general face recognition, there are billions of images in a dataset, while for neuroim-aging there are just a thousand. As initialization, CNNs that have been pretrained on a dataset can be utilized for classification of neuroimaging data to classify AD. The CNNs may then be trained again on neuros-cans using TL. Thus, researchers are able to classify AD more efficiently using pre-trained models, such as VGG (16, 19), ResNet, DenseNet (121,161,169), LeNet, and Inception as compared to building a model from scratch. Researchers have achieved the highest accuracy for AD vs. NC (approx 99.5%) in comparison to other binary (AD v/s MCI, MCI v/s NC, cMCI v/s ncMCI), ternary (AD v/s MCI v/s NC, AD vs. cMCI v/s ncMCI) and multiclass classification (NonD v/s VMild vs. Mild vs. Mod).
References
1. Shen, D., Wu, G., Suk, H.I., Deep learning in medical image analysis.
Annu. Rev. Biomed. Eng.
, 19, 221–248, 2017.
2. Taqi, A.M., Awad, A., Al-Azzo, F., Milanova, M., The impact of multi- optimizers and data augmentation on TensorFlow convolutional neural network performance, in:
2018 IEEE Conference on Multimedia Information Processing and Retrieval (MIPR)
, April 2018, IEEE, pp. 140–145.
3. Sethi, M., Ahuja, S., Rani, S., Bawa, P., Zaguia, A., Classification of Alzheimer’s disease using gaussian-based bayesian parameter optimization for deep con-volutional LSTM network.
Comput. Math. Methods Med.
, 2021, 2021.
4. Liu, M., Li, F., Yan, H., Wang, K., Ma, Y., Shen, L., Xu, M., Alzheimer’s disease neuroimaging initiative, a multi-model deep convolutional neural network for automatic hippocampus segmentation and classification in Alzheimer’s disease.
Neuroimage
, 208, 116459, 2020.
5. Ebrahimighahnavieh, M.A., Luo, S., Chiong, R., Deep learning to detect Alzheimer’s disease from neuroimaging: A systematic literature review.
Comput. Methods Programs Biomed.
, 187, 105242, 2020.
6. Kadyan, V., Bala, S., Bawa, P., Mittal, M., Developing in-vehicular noise robust children ASR system using Tandem-NN-based acoustic modelling.
Int. J. Veh. Auton. Syst.
, 15, 3-4, 296–306, 2020.
7. Suk, H.I., Lee, S.W., Shen, D., Deep sparse multi-task learning for feature selection in Alzheimer’s disease diagnosis.
Brain Struct. Funct.
, 221, 5, 2569– 2587, 2016.
8. Sethi, M., Ahuja, S., Bawa, P., Classification of Alzheimer’s disease using neu-roimaging data by convolution neural network, in:
2021 6th International Conference on Signal Processing, Computing and Control (ISPCC)
, 2021, October, IEEE, pp. 402–406.
9. Sethi, M., Ahuja, S., Kukreja, V., An empirical study for the deep learning models.
J. Phys.: Conf. Ser.
, 1950, 1, 012071, IOP Publishing, 2021.
10. Sethi, M., Ahuja, S., Rani, S., Koundal, D., Zaguia, A., Enbeyle, W., An exploration: Alzheimer’s disease classification based on convolutional neural network.
BioMed. Res. Int.
, 2022, 2022.
11. Verma, K., Bhardwaj, S., Arya, R., Islam, M.S.U., Bhushan, M., Kumar, A., Samant, P., Latest tools for data mining and machine learning.
Int. J. Innov. Technol. Exploring Eng.
, 8, 9S, 18–23, July 2019, Available:
https://doi.org/10.35940/ijitee.I1003.0789S19
.
12. Nalavade, A., Bai, A., Bhushan, M., Deep learning techniques and models for improving machine reading comprehension system.
IJAST
, 29, 04, 9692– 9710, 2020.
13. Ashraf, A., Naz, S., Shirazi, S.H., Razzak, I., Parsad, M., Deep transfer learning for Alzheimer neurological disorder detection.
Multimed. Tools Appl.
, 1–26, 2021.
14. Yang, Q. and Wu, X., 10 Challenging Problems in Data Mining Research.
Int. J. Inf. Technol. Decis. Mak.
, 5, 04, 597–604, 2006.
15. Pan, S.J. and Yang, Q., A survey on transfer learning.
IEEE Trans. Knowl. Data Eng.
, 22, 10, 1345–1359, 2009.
16. Weiss, K., Khoshgoftaar, T.M., Wang, D., A survey of transfer learning.
J. Big Data
, 3, 1, 1–40, 2016.
17. Kholiya, P.S., Kapoor, A., Rana, M., Bhushan, M., Intelligent process automation: The future of digital transformation, in:
2021 10th International Conference on System Modeling & Advancement in Research Trends (SMART)
, IEEE, pp. 185–190, 2021.
18. Wu, H., Luo, J., Lu, X., Zeng, Y., 3D transfer learning network for classifi-cation of Alzheimer’s disease with MRI.
Int. J. Mach. Learn. Cybern.
, 13, 2, 1–15, 2022.
19. Sayeedakhanum, P., Megha, B., Anita, B., A study on healthcare using data mining techniques.
J. Crit. Rev.
, 7, 19, 7877–7890, 2020.
20. Bawa, P. and Kadyan, V., Noise robust in-domain children speech enhancement for automatic Punjabi recognition system under mismatched conditions.
Appl. Acoust.
, 175, 107810, 2021.
21. Singh, V.J., Bhushan, M., Kumar, V., Bansal, K.L., Optimization of segment size assuring application perceived QoS in healthcare, in:
Proceedings of the World Congress on Engineering
, vol. 1, 2015.
22. Hu, C., Ju, R., Shen, Y., Zhou, P., Li, Q., Clinical decision support for Alzheimer’s disease based on deep learning and brain network, in:
2016 IEEE International Conference on Communications (ICC)
, May 2016, IEEE, pp. 1–6.
23. Shafi, I., Ansari, S., Din, S., Jeon, G., Paul, A., Artificial neural networks as clinical decision support systems.
Concurr. Comput.: Pract. Exp.
, 33, 22, e6342, 2021.
24. Gupta, A., Ayhan, M., Maida, A., Natural image bases to represent neuro-imaging data, in:
International Conference on Machine Learning
, May 2013, PMLR, pp. 987–994.
25. Liu, S., Liu, S., Cai, W., Pujol, S., Kikinis, R., Feng, D., Early diagnosis of Alzheimer’s disease with deep learning, in:
2014 IEEE 11th International Symposium on Biomedical Imaging (ISBI)
, April 2014, IEEE, pp. 1015–1018.
26. Suk, H.I., Lee, S.W., Shen, D., Alzheimer’s disease neuroimaging initiative, hierarchical feature representation and multimodal fusion with deep learning for AD/MCI diagnosis.
NeuroImage
, 101, 569–582, 2014.
27. Payan, A. and Montana, G., Predicting Alzheimer’s disease: A neuroimaging study with 3D convolutional neural networks. in
Proc. Int. Conf. Pattern Recogn. Appl. Meth.,
2015.
28. Gao, X.W. and Hui, R., A deep learning based approach to classification of CT brain images, in:
2016 SAI Computing Conference (SAI)
, July 2016, IEEE, pp. 28–31.
29. Ortiz, A., Munilla, J., Gorriz, J.M., Ramirez, J., Ensembles of deep learning architectures for the early diagnosis of the Alzheimer’s disease.
Int. J. Neural Syst.
, 26, 07, 1650025, 2016.
30. Cheng, D. and Liu, M., Classification of Alzheimer’s disease by cascaded con-volutional neural networks using PET images, in:
International Workshop on Machine Learning in Medical Imaging
, September 2017, Springer, Cham, pp. 106–113.
31. Luo, S., Li, X., Li, J., Automatic Alzheimer’s disease recognition from MRI data using deep learning method.
J. Appl. Math. Phys.
,
5
, 9, 1892–1898, 2017.
32. Cheng, D. and Liu, M., CNNs based multi-modality classification forAD diagnosis, in:
2017 10th International Congress on Image and Signal Processing, Biomedical Engineering and Informatics (CISP-BMEI)
, October 2017, IEEE, pp. 1–5.
33. Korolev, S., Safiullin, A., Belyaev, M., Dodonova, Y., Residual and plain con-volutional neural networks for 3D brain MRI classification, in:
2017 IEEE 14th International Symposium on Biomedical Imaging (ISBI 2017)
, April 2017, IEEE, pp. 835–838.
34. Cheng, D., Liu, M., Fu, J., Wang, Y., Classification of MR brain images by combination of multi-CNNs for AD diagnosis, in:
Ninth International Conference on Digital Image Processing (ICDIP 2017)
, July 2017, vol. 10420, International Society for Optics and Photonics, p. 1042042.
35. Lin, W., Tong, T., Gao, Q., Guo, D., Du, X., Yang, Y., Guo, G., Xiao, M., Du, M., Qu, X., Alzheimer’s disease neuroimaging initiative, convolutional neural networks-based MRI image analysis for the Alzheimer’s disease prediction from mild cognitive impairment.
Front. Neurosci.
, 12, 777, 2018.
36. Liu, M., Cheng, D., Yan, W., Alzheimer’s Disease Neuroimaging Initiative, Classification of Alzheimer’s disease by combination of convolutional and recurrent neural networks using FDG-PET images.
Front. Neuroinform.
, 12, 35, 2018.
37. Shi, J., Zheng, X., Li, Y., Zhang, Q., Ying, S., Multimodal neuroimaging feature learning with multimodal stacked deep polynomial networks for diagnosis of Alzheimer’s disease.
IEEE J. Biomed. Health Inform.
, 22, 1, 173–183, 2017.
38. Huang, Y., Xu, J., Zhou, Y., Tong, T., Zhuang, X., Alzheimer’s disease neuro-imaging initiative (ADNI, diagnosis of Alzheimer’s disease via multi-modality 3D convolutional neural network.
Front. Neurosci.
, 13, 509, 2019.
39. Feng, C., Elazab, A., Yang, P., Wang, T., Zhou, F., Hu, H., Xiao, X., Lei, B., Deep learning framework for Alzheimer’s disease diagnosis via 3D-CNN and FSBi-LSTM.
IEEE Access
, 7, 63605–63618, 2019.
40. Choi, B.K., Madusanka, N., Choi, H.K., So, J.H., Kim, C.H., Park, H.G., Bhattacharjee, S., Prakash, D., Convolutional neural network-based MR image analysis for Alzheimer’s disease classification.
Curr. Med. Imaging
, 16, 1, 27–35, 2020.
41. Feng, W., Halm-Lutterodt, N.V., Tang, H., Mecum, A., Mesregah, M.K., Ma, Y., Li, H., Zhang, F., Wu, Z., Yao, E., Guo, X., Automated MRI-based deep learning model for detection of Alzheimer’s disease process.
Int. J. Neural Syst.
, 30, 06, 2050032, 2020.
42. Pei, Z., Gou, Y., Ma, M., Guo, M., Leng, C., Chen, Y., Li, J., Alzheimer’s disease diagnosis based on long-range dependency mechanism using convolutional neural network.
Multimed. Tools Appl.
, 1–16, 2021.
43. Ajagbe, S.A., Amuda, K.A., Oladipupo, M.A., AFE, O.F., Okesola, K.I., Multi-classification of alzheimer disease on magnetic resonance images (MRI) using deep convolutional neural network (DCNN) approaches.
