Schlagwort(e):
Mobile communication systems.
;
Electronic books.
Materialart:
Online-Ressource
Seiten:
1 online resource (224 pages)
Ausgabe:
1st ed.
ISBN:
9783031023804
Serie:
Synthesis Lectures on Learning, Networks, and Algorithms Series
URL:
https://ebookcentral.proquest.com/lib/geomar/detail.action?docID=6380567
Sprache:
Englisch
Anmerkung:
Cover -- Copyright Page -- Title Page -- Contents -- Preface -- Acknowledgments -- Introduction to Edge Intelligence -- Artificial Intelligence -- Deep Learning and Deep Neural Networks -- From Deep Learning to Model Training and Inference -- Deep Learning Applications -- Popular Deep Learning Models -- Edge Computing -- Evolution of Edge Computing -- Benefits of Edge Computing -- Edge Intelligence -- Motivation and Benefits of Edge Intelligence -- Scope and Rating of Edge Intelligence -- Edge Intelligence via Model Training -- Architectures -- Centralized -- Decentralized -- Hybrid -- Key Performance Indicators -- Training Loss -- Convergence -- Privacy -- Communication Cost -- Latency -- Energy Efficiency -- Enabling Technologies -- Federated Learning -- Aggregation Frequency Control -- Gradient Compression -- DNN Splitting -- Knowledge Transfer Learning -- Gossip Training -- Hardware Acceleration -- Knolwedge Distillation for Edge-Cloud Collaboration -- Meta-Learning -- Summary -- Edge Intelligence via Federated Meta-Learning -- Introduction -- Related Work -- Preliminaries on Meta-Learning -- Federated Meta-Learning for Achieving Real-Time Edge Intelligence -- Problem Formulation -- Federated Meta-Learning (FedML) -- Performance Analysis of FedML -- Convergence Analysis -- Performance Evaluation of Fast Adaptation -- Robust Federated Meta-Learning (FedML) -- Robust Federated Meta-Learning -- Wasserstein Distance-Based Robust Federated Meta-Learning -- Robust Meta-Training across Edge Nodes -- Convergence Analysis -- Experiments -- Experimental Setting -- Evaluation of Federated Meta-Learning -- Evaluation of Robust Federated Meta-Learning -- Summary -- Edge-Cloud Collaborative Learning via Distributionally Robust Optimization -- Introduction -- Basic Setting for Collaborating Learning toward Edge Intelligence.
,
Collaborative Learning Based on Edge-Cloud Synergy of Distribution Uncertainty Sets -- Cloud Knowledge Transfer Based on Edge-Cloud Model Relation -- DRO Formulation with Two Distribution Distance-Based Uncertainty Sets -- Construction of Wasserstein Distance-Based Uncertainty Sets -- A Duality Approach to the Wasserstein Distance-Based DRO for Edge Learning -- Experimental Results -- Collaborative Learning Based on Knowledge Transfer of Conditional Prior Distribution -- Cloud-Edge Knowledge Transfer via Dirichlet Process Prior -- DRO Formulation with Conditional Prior Distribution and Edge Distribution Uncertainty Set -- A Duality Approach to The Inner Maximization Problem -- A Convex Relaxation for The Outer Minimization Problem -- Experimental Results -- Summary -- Hierarchical Mobile-Edge-Cloud Model Training with Hybrid Parallelism -- Introduction -- Background and Motivation -- Hiertrain Framework -- Problem Statement of Policy Scheduling -- Training Tasks in Hiertrain -- Training Procedure in Hiertrain -- Minimization of Training Time -- Optimization of Policy Scheduling -- Performance Evaluation -- Dataset and Models -- Experimental Setup -- Baselines -- Results -- Summary -- Edge Intelligence via Model Inference -- Architectures -- Edge-Based -- Device-Based -- Edge-Device -- Edge-Cloud -- Key Performance Indicators -- Latency -- Accuracy -- Energy -- Privacy -- Communication Overhead -- Memory Footprint -- Enabling Technologies -- Model Compression -- Model Partition -- Edge Caching -- Input Filtering -- Model Selection -- Support for Multi-Tenancy -- Application-Specific Optimization -- Summary -- On-Demand Accelerating Deep Neural Network Inference via Edge Computing -- Introduction -- Background and Motivation -- Insufficiency of Device-Only or Edge-Only DNN Inference -- DNN Partitioning and Right-Sizing Toward Edge Intelligence.
,
Framework and Design -- Framework Overview -- Edgent for Static Environments -- Edgent for Dynamic Environments -- Performance Evaluation -- Experimental Setup -- Experiments in Static Bandwidth Environment -- Experiments in Dynamic Bandwidth Environment -- Summary -- Applications, Marketplaces, and Future Directions of Edge Intelligence -- Applications of Edge Intelligence -- Video Analytics -- Cognitive Assistance -- Industrial Internet of Things (IIoT) -- Smart Home -- Precision Agriculture -- Smart Retail -- Marketplace of Edge Intelligence -- Market Overview and Landscape -- Emerging Commercial Platforms for EI -- Emerging Commercial Devices and Hardware for EI -- Future Directions on Edge Intelligence -- Building a Theoretic Foundation of Edge Intelligence (EI) -- Learning-Driven Networking, Security, and Privacy Techniques for EI -- Programming and Software Platforms for EI -- Smart Service and Incentive Models for EI -- Bibliography -- Authors' Biographies.
Permalink
