Explainable edge AI : a futuristic computing perspective

This book presents explainability in edge AI, an amalgamation of edge computing and AI. The issues of transparency, fairness, accountability, explainability, interpretability, data-fusion, and comprehensibility that are significant for edge AI are being addressed in this book through explainable models and techniques. The concept of explainable edge AI is new in front of the academic and research community, and consequently, it will undoubtedly explore multiple research dimensions. The book presents the concept of explainability in edge AI which is the amalgamation of edge computing and AI. In the futuristic computing scenario, the goal of explainable edge AI will be to execute the AI tasks and produce explainable results at the edge. First, this book explains the fundamental concepts of explainable artificial intelligence (XAI), then it describes the concept of explainable edge AI, and finally, it elaborates on the technicalities of explainability in edge AI. Owing to the quick transition in the current computing scenario and integration with the latest AI-based technologies, it is significant to facilitate people-centric computing through explainable edge AI. Explainable edge AI will facilitate enhanced prediction accuracy with the comprehensible decision and traceability of actions performed at the edge and have a significant impact on futuristic computing scenarios. This book is highly relevant to graduate/postgraduate students, academicians, researchers, engineers, professionals, and other personnel working in artificial intelligence, machine learning, and intelligent systems. Preface Contents About the Editors 1 Explainable Artificial Intelligence: Concepts and Current Progression 1 Introduction 2 Literature Review 3 Principles 4 Models 4.1 Features-Based Approaches 4.2 Global Approaches 4.3 Concept Activation Vector Approach 4.4 LIME Approach 4.5 LRP Approach 5 Current Progression 6 XAI in Medicines 6.1 Seven Pillars of XAI in Medicine 7 Applications of XAI 8 Benefits of XAI 9 Challenges in XAI 10 Future of XAI 11 Conclusion References 2 Explainable Artificial Intelligence (XAI): Understanding and Future Perspectives 1 Introduction 2 Need for Transparency and Trust in AI 3 Methods of Explainable AI 3.1 Perturbation-Based 3.2 Backpropagation- or Gradient-Based 4 XAI Application Domains 5 XAI Challenges and Future Prospective 6 Conclusion References 3 Explainable Artificial Intelligence (XAI): Conception, Visualization and Assessment Approaches Towards Amenable XAI 1 Introduction 2 Preliminaries and Definitions 3 Techniques for Explainability and Interpretability 3.1 Integrated Interpretability 3.2 Post-Hoc Methods 4 Empirical Analysis 4.1 Machine Learning 5 Conclusion References 4 Explainable AI (XAI): A Survey of Current and Future Opportunities 1 Introduction 1.1 Understanding Artificial Intelligence 1.2 Advancement of the AI 1.3 Special Considerations 1.4 How Artificial Intelligence Can Affect Human Activities? 2 Applications of AI 3 Importance of AI in Health Industry 4 Introduction to Explainable Artificial Intelligence 5 Background of AI and XAI 6 Challenges 7 Applications of XAI 8 Difference Between AI and XAI 9 Conclusion and Future Scope References 5 Recent Challenges on Edge AI with Its Application: A Brief Introduction 1 Introduction 2 Beyond Perspective with 5G 2.1 Communication and Computation Process Using Human 2.2 Applications in Scale 2.3 Edge Intelligence System and Latest Technologies Beyond 5G Networks 2.4 Technology Meets Business 3 Future Perspective on Cloud Computing 3.1 Resource Management 3.2 Energy and Operational Constraints 3.3 Security and Privacy Issues 3.4 Intermittent Connectivity 4 Evolving AI and ML 4.1 Accelerators for AI Usage 4.2 Trade off Between Accuracy and Resource Demand 4.3 Federated Learning 5 Outlook and Roadmap in Edge AI 5.1 Open Research Challenges 5.2 Safety and Privacy/Ethical Issues 5.3 Conclusion and Future Scope References 6 Explainable Artificial Intelligence in Health Care: How XAI Improves User Trust in High-Risk Decisions 1 Introduction 2 Explainable AI Principles 3 Why Is Explainable AI Important? 4 Explainable AI Example 5 Explainable Artificial Intelligence in Medical and Industrial Applications 6 XAI in Finances 7 Explainable AI in the Automotive Industry 8 Explainable Artificial Intelligence in Manufacturing 9 Achieving XAI in Health Care 10 Why XAI is Important in Health Care 11 A Case Study “Life or Death: West Nile Virus” 12 XAI—Issues and Challenges 13 Conclusion References 7 Role of Explainable Edge AI to Resolve Real Time Problem 1 Introduction 1.1 Definition of Explainable Artificial Intelligence 1.2 Edge AI Differ from Artificial Intelligence 1.3 Explainable Edge AI Can Assist Internet of Behavior Effort 1.4 Organization of the Chapter 2 Concepts of XAI 2.1 Major Principles in XAI 2.2 Working Scenario of XAI Principles 2.3 Benefits of XAI 2.4 Challenges in XAI 3 Example of XAI in Real World 3.1 XAI in Defense 3.2 XAI in Autonomous Vehicle 3.3 XAI in Fraudulent Activities 3.4 XAI in Marketing 3.5 Summary 4 Future of XAI 5 Conclusion References 8 Explainable Data Fusion on Edge: Challenges and Opportunities 1 Introduction 2 Data Fusion Architecture and Models 2.1 JDL/DGIF Model 2.2 Waterfall Model 2.3 Omnibus Model 3 Edge Intelligence 3.1 Challenges and Resolutions of Edge Intelligence 4 Explainable Artificial Intelligence 4.1 Terminology in XAI 4.2 Purpose of XAI 5 The How of Explainability in AI 6 Application of XAI for Data Fusion 7 Conclusion References 9 Trust Model Based Data Fusion in Explainable Artificial Intelligence for Edge Computing Using Secure Sequential Discriminant Auto Encoder with Lightweight Optimization Algorithm 1 Introduction 2 Background 3 System Model 3.1 Secure Sequential Discriminant Auto Encoder (SSDAE) Based Data Fusion 3.2 Secure Sequential Fuzzy Based Trust Model (SSFTM) 3.3 Data Optimization Using Genetic Swarm Lightweight Optimization Algorithm 4 Performance Analysis 5 Conclusion References 10 A Deep Learning Based Target Coverage Protocol for Edge Computing Enabled Wireless Sensor Networks 1 Introduction 2 Machine Learning in WSN 3 Network Model 4 Proposed Protocol 5 Simulation Setup and Parameters 6 Simulation Results and Discussion 6.1 Model Analysis 6.2 Explainability of the Proposed Model Using Lime 6.3 LGBM Classifier 6.4 Global Feature Importance 7 Conclusion and Future Scope References