Advances in Data Science and Analytics: Concepts and Paradigms

ISBN-13: 9781119791881 / Angielski / Twarda / 2022 / 350 str.

M. Niranjanamurthy; Hemant Kumar Gianey; Amir H. Gandomi

Spis treści

Preface xv1 Implementation Tools for Generating Statistical Consequence Using Data Visualization Techniques 1Dr. Ajay B. Gadicha, Dr. Vijay B. Gadicha, Prof. Sneha Bohra and Dr. Niranjanamurthy M.1.1 Introduction 21.2 Literature Review 41.3 Tools in Data Visualization 41.4 Methodology 141.4.1 Plotting the Data 141.4.2 Plotting the Model on Data 151.4.3 Quantifying Linear Relationships 161.4.4 Covariance vs. Correlation 171.5 Conclusion 18References 182 Decision Making and Predictive Analysis for Real Time Data 21Umesh Pratap Singh2.1 Introduction 222.2 Data Analytics 232.2.1 Descriptive Analytics 232.2.2 Diagnostic Analytics 232.2.3 Predictive Analytics 232.2.4 Prescriptive Analytics 242.3 Predictive Modeling 242.4 Categories of Predictive Models 242.5 Process of Predictive Modeling 252.5.1 Requirement Gathering 262.5.2 Data Gathering 262.5.3 Data Analysis and Massaging 262.5.4 Machine Learning Statistics 262.5.5 Predictive Modeling 262.5.6 Prediction and Decision Making 272.6 Predictive Analytics Opportunities 272.6.1 Detecting Fraud 272.6.2 Reduction of Risk 272.6.3 Marketing Campaign Optimization 282.6.4 Operation Improvement 282.6.5 Clinical Decision Support System 282.7 Classification of Predictive Analytics Models 282.7.1 Predictive Models 282.7.2 Descriptive Models 292.7.3 Decision Models 292.8 Predictive Analytics Techniques 292.8.1 Predictive Analytics Software 292.8.2 The Importance of Good Data 302.8.3 Predictive Analytics vs. Business Intelligence 302.8.4 Pricing Information 302.9 Data Analysis Tools 302.9.1 Excel 302.9.2 Tableau 312.9.3 Power BI 312.9.4 Fine Report 312.9.5 R & Python 312.10 Advantages & Disadvantages of Predictive Modeling 312.10.1 Advantages 312.10.2 Disadvantages 322.10.2.1 Data Labeling 322.10.2.2 Obtaining Massive Training Datasets 322.10.2.3 The Explainability Problem 322.10.2.4 Generalizability of Learning 332.10.2.5 Bias in Algorithms and Data 332.11 Predictive Analytics Biggest Impact 332.11.1 Predicting Demand 332.11.2 Transformation Using Technology and Process 342.11.3 Improved Pricing 342.11.4 Predictive Maintenance 352.12 Application of Predictive Analytics 352.12.1 Financial and Banking Services 352.12.2 Retail 352.12.3 Health and Insurance 362.12.4 Oil and Gas Utilities 362.12.5 Public Sector 362.13 Future Scope of Predictive Modeling 362.13.1 Technological Advancements 372.13.2 Changes in Work 372.13.3 Risk Mitigation 372.14 Conclusion 37References 383 Optimizing Water Quality with Data Analytics and Machine Learning 39Bin Liang, Zhidong Li, Hongda Tian, Shuming Liang, Yang Wang and Fang Chen3.1 Introduction 393.2 Related Work 413.3 Data Sources and Collection 423.4 Water Demand Forecasting 433.4.1 Network Flow and Zone Demand Estimation 433.4.2 Demand Forecasting 443.4.2.1 Feature Importance 453.4.2.2 Forecast Horizon 463.4.3 Performance Characterization 463.5 Re-Chlorination Optimization 493.5.1 Data 513.5.2 Water Age Estimation 523.5.2.1 Travel Time Estimation 533.5.2.2 Residential Time Estimation 543.5.3 Ammonia Prediction 543.5.4 Optimization Model Definition 573.5.5 Improvements in Customer Water Quality 593.5.6 Plant Dosing Optimization 623.6 Conclusion 63Acknowledgements 63References 634 Lip Reading Framework using Deep Learning and Machine Learning 67Hemant Kumar Gianey, Parth Khandelwal, Prakhar Goel, Rishav Maheshwari, Bhannu Galhotra and Divyanshu Pratap Singh4.1 Introduction 684.1.1 Overview 684.1.2 Motivation 684.1.3 Lip Reading System Outcomes and Deliverables 694.2 The Emergence and Definition of the Lip-Reading System 704.2.1 Background of Domain 704.2.2 Identified Problems 784.2.3 Tools and Technologies Used 784.2.4 Implementation Aspects 784.2.4.1 Data Preparation 794.3 Design and Components of Lip-Reading System 824.4 Lip Reading System Architecture 824.5 Testing 844.6 Problems Encountered During Implementation 844.6.1 Assumptions and Constraints 854.7 Conclusion 854.8 Future Work 85References 865 New Perspective to Management, Economic Growth and Debt Nexus Analysis: Evidence from Indian Economy 89Edmund Ntom Udemba, Festus Victor Bekun, Dervis Kirikkaleli and Esra Sipahi Döngül5.1 Introduction 905.2 Literature Review 925.2.1 External Debt and Economic Growth 925.2.2 Trade Openness, FDI, and Economic Growth 945.2.3 FDI and Economic Growth 945.3 Data 955.3.1 Analytical Framework and Data Description 965.3.2 Theoretical Background and Specifications 965.3.2.1 Model Specification 985.4 Methodology and Findings 995.4.1 Unit Root Testing 995.4.2 Cointegration 995.4.3 Vector Error Correction Model 1035.4.4 Long-Run Relationship Estimation 1055.4.5 Causality Test 1075.5 Conclusion and Policy Implications 108Declarations 109Availability of Data and Materials 109Competing Interests 110Funding 110Authors' Contributions 110Acknowledgments 110References 1106 Data-Driven Delay Analysis with Applications to Railway Networks 115Boyu Li, Ting Guo, Yang Wang and Fang Chen6.1 Introduction 1166.2 Related Works 1186.3 Background Knowledge 1196.3.1 Background and Problem Formulation 1206.3.1.1 Train Delay 1206.3.1.2 Delay Propagation 1216.3.2 Preliminaries 1226.3.2.1 Bayesian Inference 1236.3.2.2 Markov Property 1236.4 Delay Propagation Model 1236.4.1 Conditional Bayesian Delay Propagation 1236.4.1.1 Delay Self-Propagation 1246.4.1.2 Incremental Run-Time Delay 1256.4.1.3 Incremental Dwell Time Delay 1256.4.1.4 Accumulative Departure Delay 1266.4.2 Cross-Line Propagation, Backward Propagation and Train Connection Propagation 1276.5 Primary Delay Tracing Back 1306.5.1 Delay Candidates Selection 1306.5.2 Relation Construction 1316.5.2.1 Preceding and Following Trains 1316.5.2.2 Preceding and Connecting Trains 1316.6 Evaluation on Dwell Time Improvement Strategy 1326.7 Experiments 1356.7.1 Experiment Setting 1356.7.2 Temporal Prediction of Delay Propagation 1376.7.3 Spatial Prediction of Delay Propagation 1386.7.4 Case Study of Primary Delay Tracing Down 1396.7.5 Evaluation of Dwell Time Improvement Strategy 1406.8 Conclusion 142References 1427 Proposing a Framework to Analyze Breast Cancer in Mammogram Images Using Global Thresholding, Gray Level Co-Occurrence Matrix, and Convolutional Neural Network (CNN) 145Ms. Tanishka Dixit and Ms. Namrata Singh7.1 Introduction & Purpose of Study 1467.1.1 Segmentation 1467.1.1.1 Types of Segmentation 1477.1.2 Compression 1507.2 Literature Review & Motivation 1537.3 Proposed Work 1617.3.1 Algorithm 1617.3.2 Explanation 1627.3.3 Flowchart 1627.4 Observation Tables and Figures 1637.5 Conclusion 1767.6 Future Work 176References 1768 IoT Technologies for Smart Healthcare 181Rehab A. Rayan, Imran Zafar and Christos Tsagkaris8.1 Introduction 1828.2 Literature Review 1838.2.1 IoT-Based Smart Health 1838.2.2 Advantages of Applying IoT in Health 1868.3 Findings 1878.3.1 Significant Features and Applications of IoT in Health 1878.3.1.1 Simultaneous Monitoring and Reporting 1898.3.1.2 End-to-End Connectivity and Affordability 1908.3.1.3 Data Analysis 1908.3.1.4 Tracking, Alerts, and Remote Medical Care 1908.3.1.5 Research 1918.3.1.6 Patient-Generated Health Data (PGHD) 1918.3.1.7 Management of Chronic Diseases and Preventative Care 1918.3.1.8 Home-Based and Short-Term Care 1928.4 Case Study: CyberMed as an IoT-Based Smart Health Model 1928.5 Discussions 1938.5.1 Limitations of Adopting IoT in Health 1938.5.1.1 Data Security and Privacy 1938.5.1.2 Connectivity 1948.5.1.3 Compatibility and Data Integration 1958.5.1.4 Implementation Cost 1958.5.1.5 Complexity and Risk of Errors 1958.6 Future Insights 1968.7 Conclusions 197References 1979 Enhancement of Scalability of SVM Classifiers for Big Data 203Vijaykumar Bhajantri, Shashikumar G. Totad and Geeta R. Bharamagoudar9.1 Introduction 2049.2 Support Vector Machine 2059.2.1 Challenges 2089.3 Parallel and Distributed Mechanism 2099.3.1 Shared-Memory Parallelism 2099.4 Distributed Big Data Architecture 2109.4.1 Hadoop MapReduce 2109.4.2 Spark 2109.4.3 Akka 2119.5 Distributed High Performance Computing 2129.5.1 GASNet 2129.5.2 Charm++ 2139.6 GPU Based Parallelism 2149.6.1 Cuda 2159.6.2 OpenCL 2159.7 Parallel and Distributed SVM Algorithms 2179.7.1 Ls-svm 2189.7.2 Cascade SVM 2199.7.3 dc Svm 2209.7.4 Parallel Distributed Multiclass SVM Algorithms 2229.8 Conclusion and Future Research Directions 222References 22510 Electrical Network-Related Incident Prediction Based on Weather Factors 233Hongda Tian, Jessie Nghiem and Fang Chen10.1 Introduction 23310.2 Related Work 23510.3 Methodology 23510.3.1 Binary Classification of Incident and Normality 23510.3.2 Incident Categorization Using Natural Language Processing 23610.3.3 Classification of Multiple Types of Incidents 23610.4 Experiments 23710.4.1 Data Sets 23710.4.2 Evaluation Metrics 23910.4.3 Binary Classification 23910.4.4 Incident Categorization 24110.4.5 Multi-Class Classification 24210.5 Conclusion and Future Work 244Acknowledgements 244References 24511 Green IoT: Environment-Friendly Approach to IoT 247Abhishek Goel and Siddharth Gautam11.1 Introduction 24711.2 G-IoT (Green Internet of Things) 24911.3 Layered Architecture of G-IoT 25111.3.1 Data Center/Cloud 25211.3.2 Data Analytics and Control Applications It 25211.3.3 Data Aggregation and Storage 25311.3.4 Edge Computing 25311.3.5 Communication and Processing Unit 25411.4 Techniques for Implementation of G-IoT 25711.5 Power Saving Methods Based on Components 26611.6 Applications of G-IoT 26611.7 Challenges and Future Scope 26911.8 Case Study 26911.9 Conclusion 270References 27112 Big-Data Analytics: A New Paradigm Shift in Micro Finance Industry 275Vinay Pal Singh, Rohit Bansal and Ram Singh12.1 Introduction 27612.2 Reality of Area and Transcendent Difficulties 27612.2.1 Probable Overlending 27812.2.2 Information Imbalance 27812.2.3 Retreating Not-for-Profit Sector 27812.2.4 Neighbourhood Pressure 27912.3 Data Analytics in Microfinance 28012.3.1 Types of Data Analytics Used in Microfinance 28012.3.2 Use of Big Data in Microfinance Industry 28112.3.3 Risk and Data Based Credit Decisions 28212.3.4 Product Development and Selection 28312.3.5 Product or Service Positioning 28312.3.6 M-Commerce and E-Payments 28312.3.7 Making Reliable Credit Decisions 28412.3.8 Big Data-Driven Model Promises Psychometric Evaluations 28412.3.9 Product Build-Up, Service Positioning, and Offering 28412.4 Opportunities and Risks in Using Data Analytics 28412.5 Risk in Utilizing Big Data 28712.6 Conclusion 290References 29013 Big Data Storage and Analysis 293Namrata Dhanda13.1 Introduction 29313.1.1 6 V's of Big Data 29413.1.2 Types of Data 29513.1.3 Issues in Handling Big Data 29713.2 Hadoop as a Solution to Challenges of Big Data 29713.2.1 The Hadoop Ecosystem 29813.2.2 Rack Awareness Policy in HDFS 30713.3 In-Memory Storage and NoSQL 30813.3.1 Key-Value Data Stores 30913.3.2 Document Stores 30913.3.3 Wide Column Stores 31013.3.4 Graph Stores 31013.3.5 Multi-Modal Databases 31013.4 Advantages of NoSQL Database 31013.5 Conclusion 311References 31114 A Framework for Analysing Social Media and Digital Data by Applying Machine Learning Techniques for Pandemic Management 313Mutyala Sridevi14.1 Introduction 31414.2 Literature Review 31414.3 Understanding Pandemic Analogous to a Disaster 31714.4 Application of Machine Learning Techniques at Various Phases of Pandemic Management 31814.4.1 Mitigation Phase 31914.4.2 Preparedness Phase 32014.4.3 Response Phase 32114.4.4 Recovery Phase 32114.5 Generalized Framework to Apply Machine Learning Techniques for Pandemic Management 32214.6 Conclusion 324References 324About the Editors 327Index 329

O autorze

M. Niranjanamurthy, PhD, is an assistant professor in the Department of Computer Applications, M. S. Ramaiah Institute of Technology, Bangalore, Karnataka, India. He earned his PhD in computer science at JJTU. He has over 13 years of teaching experience and two years of industry experience as a software engineer. He has published four books and 85 papers in technical journals and conferences. He has six patents to his credit and has won numerous awards.Hemant Kumar Gianey, PhD, is a senior assistant professor in the Computer Science Department at Vellore Institute of Technology, AP, India. He also worked at Thapar Institute of Engineering and Technology, Patiala, Punjab, India and worked as a post-doctoral researcher in the Computer Science and Engineering Department at National Cheng Kung University in Taiwan. He has over 15 years of teaching and industry experience. He has conducted many workshops and has been a guest speaker in various universities. He has also published many research papers on in scientific and technical journals.Amir H. Gandomi, PhD, is a professor of data science in the Department of Engineering and Information Technology, University of Technology Sydney. Before joining UTS, he was an assistant professor at the School of Business, Stevens Institute of Technology, NJ, and a distinguished research fellow at BEACON Center, Michigan State University. He has published over 150 journal papers and four books and collectively has been cited more than 14,000 times. He has been named as one of the world's most influential scientific minds and a Highly Cited Researcher (top 1%) for three consecutive years, from 2017 to 2019. He has also served as associate editor, editor, and guest editor in several prestigious journals and has delivered several keynote talks. He is also part of a NASA technology cluster on Big Data, Artificial Intelligence, and Machine Learning.