Parallel and Distributed Computing, Applications and Technologies: 21st International Conference, PDCAT 2020, Shenzhen, China, December 28–30, 2020, ... (Lecture Notes in Computer Science, 12606)

✍ Scribed by Yong Zhang (editor), Yicheng Xu (editor), Hui Tian (editor)

• Publisher: Springer
• Year: 2021
• Tongue: English
• Leaves: 414
• Category: Library

✦ Synopsis

This book constitutes the proceedings of the 21^st International Conference on Parallel and Distributed Computing, Applications, and Technologies, PDCAT 2020, which took place in Shenzhen, China, during December 28-30, 2020.

The 34 full papers included in this volume were carefully reviewed and selected from 109 submissions. They deal with parallel and distributed computing of networking and architectures, software systems and technologies, algorithms and applications, and security and privacy.

✦ Table of Contents

Preface
Organization
Contents
Blood Leukocyte Object Detection According to Model Parameter-Transfer and Deformable Convolution
1 Introduction
2 Related Work
2.1 Deformable Convolution
2.2 Region of Interest Pooling
2.3 Parameter-Based Transfer Learning
3 Formulation
3.1 Main Idea
3.2 Deformable Convolution Network Structure
3.3 Deformable Position-Sensitive RoI Pooling
3.4 Model Framework for Blood Leukocyte Object Detection on the Basis of Model Parameter-Transfer and Deformable Convolution
3.5 Algorithm
4 Experiments
4.1 Datasets and Experimental Environment
4.2 Experimental Results and Analysis
5 Conclusion and Discussion
References
Heterogeneous Software Effort Estimation via Cascaded Adversarial Auto-Encoder
1 Introduction
1.1 Motivation
1.2 Contribution
1.3 Organization
2 Related Work
2.1 Homogeneous Effort Estimation
2.2 Heterogeneous Related Work in Software Effort Estimation
3 The Proposed Approach
3.1 Overview of Our Approach
3.2 Cascaded Adversarial Auto-Encoder
3.3 Feature Enhancement and Effort Estimation Procedures
4 Experimental Design
4.1 Datasets
4.2 Evaluation Measures
4.3 Compared Methods
4.4 Experimental Settings
5 Experimental Results and Conclusions
References
A Novel Distributed Reinforcement Learning Method for Classical Chinese Poetry Generation
1 Introduction
2 Related Work
2.1 Poetry Generation
2.2 Reinforcement Learning
3 The Proposed Method
3.1 Stage One
3.2 Stage Two
3.3 Algorithm
4 Experiment
4.1 Data Set
4.2 Experimental Settings
4.3 Baselines
4.4 Experimental Results and Analysis
4.5 Efficiency Test
5 Conclusion
A More Examples of Our Method
References
Memory Access Optimization of High-Order CFD Stencil Computations on GPU
1 Introduction
2 Using Registers to Optimize Stencil Computations
2.1 Stencil Computations in HNSC
2.2 CUDA Warp Shuffle Functions
2.3 Using Warp Shuffle Functions in Stencil Computations
3 Using Shared Memory to Optimize Stencil Computations
3.1 The GPU Shared Memory and Its Characteristics
3.2 Shared Memory Based Implementation
4 Overall Performance
5 Conclusion
References
The Dataflow Runtime Environment of DFC
1 Introduction
2 Overview of DFC
2.1 DF Function
2.2 Active Data
2.3 Program Framework in DFC
3 DFC Runtime
3.1 DAG Management
3.2 Tasks Trigger and Execution
3.3 Distinguish Data of Different Passes
4 Evaluation
4.1 Experimental Environment
4.2 Brief of DSPatch
4.3 Experiment Results
5 Conclusion and Future Works
References
Adaptive Tensor-Train Decomposition for Neural Network Compression
1 Introduction
2 Related Work
3 The Proposed Method
3.1 Compress Traditional Network by TT Decomposition
3.2 Adaptive Tensor-Train Decomposition
3.3 Depthwise Separable Convolution Based on ATTD
3.4 Improvement of Inference Speed and Optimization of GPU
4 Experiments
4.1 Datasets
4.2 Implementation Details
4.3 Effectiveness of Proposed Method
5 Conclusion
References
Development of a UAV Path Planning Approach for Multi-building Inspection with Minimal Cost
1 Introduction
2 Related Work
3 Path Planning Algorithm
3.1 Description of the Algorithm
3.2 Cost Functions
3.3 Implementation of the Algorithm
4 Simulation Results
5 Conclusion
References
Construction of Completely Independent Spanning Tree Based on Vertex Degree
1 Introduction
2 Preliminaries
3 Number of CISTs for (G) = n(G) - 2
4 Number of CISTs for (G) = n(G) - 3
5 Conclusions
References
Distributed Algorithm for Truss Maintenance in Dynamic Graphs
1 Introduction
2 Related Work
3 Preliminaries
4 Distributed Truss Decomposition
5 Distributed Truss Maintenance
6 Experiments
7 Conclusions
References
Short-Term Load Forecasting Based on CNN-BiLSTM with Bayesian Optimization and Attention Mechanism
1 Introduction
2 Methodologies
2.1 Convolutional Neural Networks (CNN)
2.2 Long Short-Term Memory (LSTM)
2.3 Bidirectional LSTM (BiLSTM)
2.4 Attention Mechanism (AM)
2.5 Bayesian Optimization (BO)
3 Evaluation Indexes
4 Experiments
4.1 Data Description
4.2 Results and Discussion
5 Conclusions
References
On the Non-ergodic Convergence Rate of the Directed Nonsmooth Composite Optimization
1 Introduction
2 DSPG Development
2.1 DSPG Algorithm
2.2 Theoretical Aspects of DSPG
3 Proof Sketch
4 Experiments
4.1 DSPG vs. SGD-PS Under the Strongly Convex Condition
4.2 DSPG vs. SGD-PS Under the Generally Convex Condition
5 Conclusion
References
6D Pose Estimation Based on the Adaptive Weight of RGB-D Feature
1 Introduction
2 Related Work
3 Model
3.1 Artificial Selection
3.2 Variant of Autoencoder (AE)
3.3 6D Pose Estimation
4 Experiments
4.1 Datasets
4.2 Metrics
4.3 Implementation Details
4.4 Result on LINEMOD Dataset
4.5 Result on Our Dataset
5 Conclusion
References
Blockchain-Based Secure Outsourcing of Fully Homomorphic Encryption Using Hidden Ideal Lattice
1 Introduction
2 Related Work
3 Notations and Background
3.1 Fully Homomorphic Encryption Using Hidden Ideal Lattice
3.2 Securely Outsourcing the Extended Euclidean Algorithm
4 The Framework of Blockchain-Based Computation Outsourcing
5 Polynomial Multiplication and Modular Exponentiation Secure Outsourcing Algorithm
5.1 Secure Outsourcing Algorithm of Polynomial Multiplication
5.2 Secure Outsourcing Algorithm of Modular Exponentiation
6 Performance Evaluation
7 Conclusion
References
Multiple Projections Learning for Dimensional Reduction
1 Introduction
2 Related Work
2.1 Locality Preserving Projection
2.2 Graph Regularization Technique
3 Proposed method
3.1 Optimization
4 Experiment
4.1 Data Sets and Experiment Setup
4.2 Experiment Result and Analysis
5 Conclusions
References
Preventing DDoS Attacks on Bitcoin Memory Pool by the Dynamic Fee Threshold Mechanism
1 Introduction
2 Background
2.1 Bitcoin
2.2 Related Work
3 DDoS Attack on Bitcoin Memory Pool
3.1 Feasibility
3.2 The Effects of Attack
4 Countering DDoS Attack on Mempool
4.1 Dynamic Fee Threshold Mechanism
4.2 Analysis of Our Method
5 Experiments
5.1 Experiment Setting
5.2 Results Analysis
6 Conclusions
References
The Compiler of DFC: A Source Code Converter that Transform the Dataflow Code to the Multi-threaded C Code
1 Introduction
2 Brief of DataFlow C
2.1 DF Function
2.2 DFC Runtime
3 The Compiler for DFC
3.1 The Source and Target of Compiling
3.2 Working Principle
3.3 Lexical Analysis
3.4 Parsing
3.5 AST Transformation
4 Analysis and Evaluation
4.1 Experimental Environment
4.2 Memory Footprint Analysis
4.3 Compiling Time
5 Conclusion and Future Works
References
Online Learning-Based Co-task Dispatching with Function Configuration in Edge Computing
1 Introduction
2 System Model and Problem Formulation
2.1 System Model
2.2 Problem Formulation
3 Online Algorithms
3.1 General Ideas
3.2 Function Configuration
3.3 Task Dispatching
4 Implementation
5 Performance Evaluation
5.1 Evaluation Settings
5.2 Large-Scale Simulation Results
6 Related Work
6.1 Task Dispatching and Function Configuration
6.2 Scheduling with Reinforcement Learning
7 Conclusion
References
System-Level FPGA Routing for Logic Verification with Time-Division Multiplexing
1 Introduction
2 Preliminaries
2.1 TDM Raito
2.2 Problem Statement
3 Our Algorithm
3.1 Edge Weight Setting and Problem Reformulation
3.2 MST-Based Net Routing
3.3 Routing Result Generation
3.4 Ratio Assignment
4 Experimental Result
5 Conclusion
References
Protein Interresidue Contact Prediction Based on Deep Learning and Massive Features from Multi-sequence Alignment
1 Introduction
2 Materials and Methods
2.1 Contact Definition
2.2 Number of Effective Sequences
2.3 Input Features
2.4 Data Sets and Model Training
2.5 Criteria for Evaluation
3 Results and Analysis
3.1 Performance on Model Reliability
3.2 Performance on Precisions of Top L Predictions
4 Conclusions
References
See Fine Color from the Rough Black-and-White
1 Introduction
2 Related Work
3 Method
3.1 Encoder and Decoder
3.2 Muti-scale Residual Module
3.3 Discriminator
3.4 Loss Function
4 Experiment
4.1 Dataset
4.2 Experimental Results Analysis
5 Conclusion
References
Data Aggregation Aware Routing for Distributed Training
1 Introduction
2 System Model
2.1 Pulling Parameters
2.2 Pushing Parameters
2.3 Routing Constraints
2.4 Cache Constraints
2.5 Problem Formulation
3 Algorithm
3.1 Pulling Routing Method
3.2 Pushing Routing Method
3.3 Routing Method for Distributed Training
4 Numerical Results
5 Conclusion
References
A New Integer Programming Model for the File Transfer Scheduling Problem
1 Introduction
2 Existing Integer Programming Models
2.1 Problem Formulation
2.2 An Existing Integer Programming Model
2.3 An Extension Integer Programming Model of M1
3 New Time-Index Integer Programming Model
3.1 A New Time-Index Formulation for FTSP
3.2 An Enhanced New Time-Index Formulation for FTSP
3.3 Generation of the Upper Bound
4 Experimental Results
4.1 Instances Generation
4.2 Performance Comparison Between Different Models
4.3 Comparison of Variables and Constraints Between Different Models
5 Conclusion
References
Approximation Algorithms for the General Cluster Routing Problem
1 Introduction
2 Preliminary
2.1 Christofides' Algorithm
2.2 The Traveling Salesman Path Problem (Path TSP)
3 The General Cluster Routing Problem with Starting and Ending Vertices
4 The General Cluster Routing Problem Without Starting and Ending Vertices
5 Numerical Experiments
6 Conclusion
References
Maximizing Group Coverage in Social Networks
1 Introduction
1.1 Related Works
1.2 Contributions
1.3 Organizatioin
2 Maximizing Group Coverage Algorithm
3 Application of MGC in Social Networks
4 Numerical Experiments
5 Conclusion
References
LightLayers: Parameter Efficient Dense and Convolutional Layers for Image Classification
1 Introduction
2 Related Work
3 Methodology
3.1 Description of Convolution Layers
4 Experimental Setup
4.1 Implementation Details
4.2 Datasets
5 Results
6 Ablation Study
7 Conclusion
References
The Hybrid Navigation Method in Face of Dynamic Obstacles
1 Introduction
2 Relate Work
3 Method
4 Experiments
5 Conclusion
References
A Relaxed Balanced Lock-Free Binary Search Tree
1 Introduction
2 Related Work
3 Algorithm Description
3.1 Overview
3.2 Detailed Algorithm
3.3 Correctness
4 Experimental Results
5 Conclusion and Future Work
References
A Dynamic Parameter Tuning Method for High Performance SpMM
1 Introduction
1.1 Background and Objective
2 Related Work
3 Dynamic Parameter Tuning Method for High Performance SpMM
3.1 Concepts of the Proposed Method
3.2 Dynamic Parameter Tuning Method
4 Evaluation
4.1 Experimental Environment
4.2 Experimental Results and Discussion
4.3 Cost of the Proposed Method
5 Conclusions
References
Data Caching Based Transfer Optimization in Large Scale Networks
1 Introduction
2 Problem Statement
3 Homogeneous Model Analysis
4 Submodular Model Analysis
5 Conclusion and Future Work
References
Second-Order Convolutional Neural Network Based on Cholesky Compression Strategy
1 Introduction
2 Related Work
3 Methodology
3.1 Basic SPDNet Model
3.2 The Proposed Second-Order CNN Model
4 Experimental Results and Analyses
4.1 Datasets
4.2 Experimental Setup
4.3 The Used Models for Comparisons
4.4 Analysis of Results
5 Conclusions
References
Submodular Maximization with Bounded Marginal Values
1 Introduction
1.1 Related Work
2 Preliminaries
3 Algorithm
4 Theoretical Analysis
5 Conclusion
References
A Streaming Model for Monotone Lattice Submodular Maximization with a Cardinality Constraint
1 Introduction
2 Preliminaries
3 A Streaming Algorithm with Known OPT
4 Two-Pass Streaming Algorithm
5 Sieve-Steaming Algorithm for Lattice Submodular Maximization
6 Conclusion
References
The Prize-Collecting k-Steiner Tree Problem
1 Introduction
2 Preliminaries
3 The First Algorithm
4 The Improved Algorithm
5 Discussions
References
Optimal Algorithm of Isolated Toughness for Interval Graphs
1 Introduction
2 Preliminaries
3 Isolated Toughness for Interval Graphs
References
Graceful Performance Degradation in Apache Storm
1 Introduction
2 Motivation
3 Design Principals
3.1 System Design Consideration
3.2 Solution Overview
3.3 Performance Model
3.4 Prediction Module
3.5 Anti-saturation
3.6 Optimization Process
4 Experimental Evaluation
4.1 Experimental Setup and Benchmark Suite
4.2 Result Summary
5 Related Work
6 Conclusion
References
Author Index