Deep Learning : Theory, Architectures and Applications in Speech, Image and Language Processing

This book is a detailed reference guide on deep learning and its applications. It aims to provide a basic understanding of deep learning and its different architectures that are applied to process images, speech, and natural language. It explains basic concepts and many modern use cases through fifteen chapters contributed by computer science academics and researchers. By the end of the book, the reader will become familiar with different deep learning approaches and models, and understand how to implement various deep learning algorithms using multiple frameworks and libraries. This book is divided into three parts. The first part explains the basic operating understanding, history, evolution, and challenges associated with deep learning. The basic concepts of mathematics and the hardware requirements for deep learning implementation, and some of its popular frameworks for medical applications are also covered. The second part is dedicated to sentiment analysis using deep learning and machine learning techniques. This book section covers the experimentation and application of deep learning techniques and architectures in real-world applications. It details the salient approaches, issues, and challenges in building ethically aligned machines. An approach inspired by traditional Eastern thought and wisdom is also presented. The final part covers artificial intelligence approaches used to explain the machine learning models that enhance transparency for the benefit of users. A review and detailed description of the use of knowledge graphs in generating explanations for black-box recommender systems and a review of ethical system design and a model for sustainable education is included in this section. An additional chapter demonstrates how a semi-supervised machine learning technique can be used for cryptocurrency portfolio management. The book is a timely reference for academicians, professionals, researchers and students at engineering and medical institutions working on artificial intelligence applications. Cover 1 Title 2 Copyright 3 End User License Agreement 4 Contents 6 Foreword 12 Preface 13 List of Contributors 15 Deep Learning: History and Evolution 17 Application of Artificial Intelligence in Medical Imaging 35 Sampurna Panda1, Rakesh Kumar Dhaka1 and Babita Panda2,* 35 INTRODUCTION 35 MACHINE-LEARNING 36 Supervised Learning 38 Unsupervised Learning 38 Semi-supervised Learning 39 Active Learning 39 Reinforcement Learning 39 Evolutionary Learning 40 Introduction to Deep Learning 40 APPLICATION OF ML IN MEDICAL IMAGING 42 DEEP LEARNING IN MEDICAL IMAGING 43 Image Classification 44 Object Classification 44 Organ or Region Detection 44 Data Mining 44 The Sign-up Process 45 Other Imaging Applications 45 CONCLUSION 46 CONSENT FOR PUBLICATION 47 CONFLICT OF INTEREST 47 ACKNOWLEDGEMENT 47 REFERENCES 47 Classification Tool to Predict the Presence of Colon Cancer Using Histopathology Images 49 Saleena Thorayanpilackal Sulaiman1,*, Muhamed Ilyas Poovankavil2 and Abdul Jabbar Perumbalath3 49 INTRODUCTION 49 METHODS AND PREPARATION 52 Dataset Preparation 52 Related Works 53 METHODOLOGY 55 Convolutional Neural Network (CNN) 55 ResNet50 57 RESULTS 58 CONCLUSION 58 CONSENT FOR PUBLICATION 58 CONFLICT OF INTEREST 59 ACKNOWLEDGEMENT 59 REFERENCES 59 Deep Learning For Lung Cancer Detection 63 Sushila Ratre1,*, Nehha Seetharaman1 and Aqib Ali Sayed1 63 INTRODUCTION 63 RELATED WORKS 65 METHODOLOGY 67 VGG16 ARCHITECTURE 70 RESNET50 ARCHITECTURE 71 FLOWCHART OF THE METHODOLOGY 72 EXPERIMENTAL RESULTS 72 CONCLUDING REMARKS 74 ACKNOWLEDGEMENTS 74 REFERENCES 74 Exploration of Medical Image Super-Resolution in terms of Features and Adaptive Optimization 76 Jayalakshmi Ramachandran Nair1,*, Sumathy Pichai Pillai2 and Rajkumar Narayanan3 76 INTRODUCTION 76 LITERATURE REVIEW 77 METHODOLOGIES 78 Pre-Upsampling Super Resolution 78 Very Deep Super-Resolution Models 80 Post Upsampling Super Resolution 80 Residual Networks 81 Multi-stage Residual Networks (MDSR) 81 Balanced Two-Stage Residual Networks 82 Recursive Networks 82 Deep Recursive Convolution Network (DRCN) 83 Progressive Reconstruction Networks 83 Attention-Based Network 84 Pixel Loss 84 Perceptual Loss 84 Adversarial Loss 84 SYSTEM TOOLS 85 FINDINGS 85 CONCLUSION 85 ACKNOWLEDGEMENTS 85 REFERENCES 86 Analyzing the Performances of Different ML Algorithms on the WBCD Dataset 89 Trupthi Muralidharr1,*, Prajwal Sethu Madhav1, Priyanka Prashanth Kumar1 and Harshawardhan Tiwari1 89 INTRODUCTION 89 LITERATURE REVIEW 90 DATASET DESCRIPTION 92 PRE-PROCESSING OF DATA 92 Exploratory Data Analysis(EDA) 92 Model Accuracy: Receiver Operating Characteristic (ROC) curve: 102 RESULTS 104 CONCLUSION 104 ACKNOWLEDGEMENTS 104 REFERENCES 104 Application and Evaluation of Machine LearningAlgorithms in Classifying Cardiotocography(CTG) Signals 106 Deep SLRT: The Development of Deep Learning based Multilingual and Multimodal Sign Language Recognition and Translation Framework 119 Natarajan Balasubramanian1 and Elakkiya Rajasekar1,* 119 INTRODUCTION 119 RELATED WORKS 122 Subunit Modelling and Extraction of Manual Features and Non-manual Features 124 Challenges and Deep Learning Methods for SLRT Research 128 THE PROPOSED MODEL 131 Algorithm: 2 NMT-GAN based Deep SLRT Video Generation (Backward) 133 Training Details 133 EXPERIMENTAL RESULTS 134 CONCLUSION 138 ACKNOWLEDGEMENTS 139 REFERENCES 139 Hybrid Convolutional Recurrent Neural Network for Isolated Indian Sign Language Recognition 145 Rajasekar Elakkiya1, Archana Mathiazhagan1 and Elakkiya Rajalakshmi1,* 145 INTRODUCTION 145 RELATED WORK 147 METHODOLOGY 148 Proposed H-CRNN Framework 148 Data Acquisition, Preprocessing, and Augmentation 150 Proposed H-CRNN Architecture 152 Experiments and Results 153 CONCLUSION AND FUTURE WORK 158 ACKNOWLEDGEMENTS 159 REFERENCES 159 A Proposal of an Android Mobile Application for Senior Citizen Community with Multi-lingual Sentiment Analysis Chatbot 162 Harshee Pitroda1,*, Manisha Tiwari1 and Ishani Saha1 162 INTRODUCTION 163 LITERATURE REVIEW 165 Twitter data 166 PROPOSED FRAMEWORK 167 IMPLEMENTATION OVERVIEW 170 Exploratory Data Analysis (EDA) 170 Feature Extraction 170 Classification 171 Support Vector Machine 171 Decision Tree 172 Random Forest 173 Implementation 173 Pickling the Model 173 Translation 173 Integrating with the Android App 174 Code Snippets 174 Support Vector Machine 174 Decision Tree 175 Random Forest 175 RESULTS AND CONCLUSION 176 Results 176 Feature Extraction 176 Classification 176 CONCLUSION 180 CONSENT FOR PUBLICATION 181 CONFLICT OF INTEREST 181 ACKNOWLEDGEMENT 181 REFERENCES 181 Technology Inspired-Elaborative Education Model (TI-EEM): A futuristic need for a Sustainable Education Ecosystem 183 Anil Verma1, Aman Singh1,*, Divya Anand1 and Rishika Vij2 183 INTRODUCTION 183 BACKGROUND 186 METHODOLOGY 188 RESULT AND DISCUSSION 191 CONCLUSION 196 CONSENT FOR PUBLICATION 197 CONFLICT OF INTEREST 197 ACKNOWLEDGEMENT 197 REFERENCES 197 Knowledge Graphs for Explaination of Black-Box Recommender System 199 Mayank Gupta1 and Poonam Saini1,* 199 INTRODUCTION 199 Introduction to Recommender System 200 Introduction to Knowledge Graphs 200 RECOMMENDER SYSTEMS 202 Types of Recommender Systems 202 KNOWLEDGE GRAPHS 204 Knowledge Graphs for Providing Recommendations 206 Knowledge Graphs for Generating Explanations 206 GENERATING EXPLANATIONS FOR BLACK-BOX RECOMME-NDER SYSTEMS 207 PROPOSED CASE STUDY 211 MovieLens Dataset 212 Modules 213 Knowledge Graph Generation 213 The Proposed Approach for Case Study 214 Results 215 Graph Visualisation 217 CONCLUSION 218 REFERENCES 219 Universal Price Tag Reader for Retail Supermarket 222 Jay Prajapati1,* and Siba Panda1 222 INTRODUCTION 222 LITERATURE REVIEW 223 METHODOLOGY 225 Image Pre-processing and Cropping 225 Optical Character Recognition 229 Price of the product 230 Name of the product 231 Discounted Price 232 RESULTS AND FUTURE SCOPE 233 CONCLUDING REMARKS 234 ACKNOWLEDGEMENTS 234 REFERENCES 234 The Value Alignment Problem: Building Ethically Aligned Machines 236 Sukrati Chaturvedi1,*, Chellapilla Vasantha Lakshmi1 and Patvardhan Chellapilla1 236 INTRODUCTION 236 Value Alignment Problem 237 Approaches for Solving AI-VAP 239 Top-Down Approach 239 Limitations, Issues, and Challenges of Extant Approaches 243 Eastern Perspectives of Intelligence for Solving AI-VAP 245 Proposed Approach 246 CONCLUSION 247 REFERENCES 247 Cryptocurrency Portfolio Management Using Reinforcement Learning 250 Vatsal Khandor1,*, Sanay Shah1, Parth Kalkotwar1, Saurav Tiwari1 and Sindhu Nair1 250 INTRODUCTION 250 RELATED WORK 252 DATASET PRE-PROCESSING 255 Simple Moving Average 256 Moving Average Convergence/Divergence 257 Parabolic Stop and Reverse 257 Relative Strength Index 257 MODELING AND EVALUATION 259 Convolutional Neural Networks (CNN) 259 Dense Neural Network Model 260 CONCLUSION AND FUTURE SCOPE 263 REFERENCES 263 Subject Index 265 Back Cover 270