Neural Computing for Advanced Applications: Second International Conference, NCAA 2021, Guangzhou, China, August 27-30, 2021, Proceedings (Communications in Computer and Information Science)

✍ Scribed by Haijun Zhang (editor), Zhi Yang (editor), Zhao Zhang (editor), Zhou Wu (editor), Tianyong Hao (editor)

• Publisher: Springer
• Year: 2021
• Tongue: English
• Leaves: 774
• Category: Library

✦ Synopsis

This book presents refereed proceedings of the Second International Conference Neural Computing for Advanced Applications, NCAA 2021, held in Guangzhou, China, in August, 2021.
The 54 full papers papers were thorougly reviewed and selected from a total of 144 qualified submissions. The papers are organized in topical sections on neural network theory, cognitive sciences, neuro-system hardware implementations, and NN-based engineering applications; machine learning, data mining, data security and privacy protection, and data-driven applications; neural computing-based fault diagnosis, fault forecasting, prognostic management, and system modeling; computational intelligence, nature-inspired optimizers, and their engineering applications; fuzzy logic, neuro-fuzzy systems, decision making, and their applications in management sciences; control systems, network synchronization, system integration, and industrial artificial intelligence; computer vision, image processing, and their industrial applications; cloud/edge/fog computing, the Internet of Things/Vehicles(IoT/IoV), and their system optimization; spreading dynamics, forecasting, and other intelligent techniques against coronavirus disease (COVID-19).

✦ Table of Contents

Preface
Organization
Contents
Neural Network Theory, Cognitive Sciences, Neuro-System Hardware Implementations, and NN-Based Engineering Applications
An Optimization Method to Boost the Resilience of Power Networks with High Penetration of Renewable Energies
1 Introduction
1.1 Backgrounds
1.2 An Insight to Enhance the Grid Resilience
2 Establishment of the Optimization Model
2.1 Disposal of the Uncertainty of Renewable Energy Power
2.2 Optimization Variable
2.3 Objective Function
2.4 Constraints
3 Case Study
3.1 Case Description
3.2 Analysis of Results
4 Conclusion
Appendix
References
Systematic Analysis of Joint Entity and Relation Extraction Models in Identifying Overlapping Relations
1 Introduction
2 Related Work
3 Joint Extraction Model Comparison
3.1 Model Differences
3.2 Feature Separation Strategy
3.3 Feature Fusion Strategy
4 Results and Analysis
4.1 Datasets
4.2 Evaluation
4.3 Results
4.4 Analysis
5 Conclusions
References
Abnormality Detection and Identification Algorithm for High-Speed Freight Train Body
1 Introduction
2 Characterization of Examined Objects and Train Set
3 Abnormality Detection and Identification Based on YOLOv4
4 The Improved-YOLOv4 Model
4.1 Data Augmentation
4.2 Negative Sample Mechanism
4.3 SECSPDarknet-53
4.4 Cascade PConv Module
4.5 Integrated Batch Normalization
4.6 The Framework of Improved-YOLOv4
5 Experiment and Analysis
6 Conclusion
References
Pheromone Based Independent Reinforcement Learning for Multiagent Navigation
1 Introduction
2 Background
2.1 Multiagent Systems (MAS) and Reinforcement Learning (RL)
2.2 The Mechanism of Stigmergy
3 Method
3.1 Dueling Double Deep Q-Network with Prioritized Replay
3.2 Digital Pheromones Coordination Mechanism
4 Experiments
4.1 Minefield Navigation Environment (MNE)
4.2 Effectiveness of PCDQN
5 Conclusion
References
A Deep Q-Learning Network Based Reinforcement Strategy for Smart City Taxi Cruising
1 Introduction
2 Problem Description
2.1 Modeling
2.2 Brief of Deep Reinforcement Learning
3 Design of DQN
3.1 Network Expressed Strategy
3.2 Procedure
4 Experiments and Results
5 Conclusions and Future Work
References
Weighted Average Consensus in Directed Networks of Multi-agents with Time-Varying Delay
1 Introduction
2 Problem Statement
3 Main Results
4 Simulation
5 Conclusion
References
An Improved Echo State Network Model for Spatial-Temporal Energy Consumption Prediction in Public Buildings
1 Introduction
2 Structure of Classical ESN
3 Chain-Structure Echo State Network
4 Experiment Design
4.1 Datasets and Model Preparation
4.2 Spatio-Temporal Forecasting of Hourly Building Energy Consumption
4.3 End-to-End Experiments on Buildings Using CESN Model
5 Experiment Results
5.1 Experimental Results of Spatio-Temporal Prediction
5.2 Experimental Results of End-to-End Prediction
6 Conclusion
References
Modeling Data Center Networks with Message Passing Neural Network and Multi-task Learning
1 Introduction
2 Related Work
2.1 Network Modeling
2.2 Routing Optimization
3 Background
3.1 Problem Setup
3.2 Overview of Message Passing Neural Network
3.3 State of the Art Method: RouteNet
4 Methods
4.1 The Extended Multi-output Architecture
4.2 Loss Function Design
4.3 Sample Generation
5 Experiments
5.1 Experiment Setup
5.2 Experiment Results
6 Conclusion
References
Machine Learning, Data Mining, Data Security and Privacy Protection, and Data-Driven Applications
A Computational Model Based on Neural Network of Visual Cortex with Conceptors for Image Classification
1 Introduction
2 Methods
2.1 Spiking Neuron Model
2.2 Conceptors
3 Network Structure
3.1 Visual Cortex (V1)
3.2 The Orientation Layer (V2)
3.3 Decision Output Layer
4 Results
4.1 The MNIST Database
4.2 The ORL Face Database
4.3 The CASIA-3D FaceV1 Database
5 Conclusion
References
Smoothed Multi-view Subspace Clustering
1 Introduction
2 Preliminaries and Related Work
2.1 Graph Filtering
2.2 Multi-view Subspace Clustering
3 Proposed Methodology
3.1 Smoothed Multi-view Subspace Clustering
4 Multi-view Experiments
4.1 Dataset
4.2 Comparison Methods
4.3 Experimental Setup
4.4 Results
4.5 Parameter Analysis
5 Conclusion
References
Sample Reduction Using 1-Norm Twin Bounded Support Vector Machine
1 Introduction
2 1-TBSVM
2.1 Formulations
2.2 Solutions and Property Analysis
3 Numerical Experiments
3.1 Artificial Dataset
3.2 UCI Datasets
4 Conclusion
References
Spreading Dynamics Analysis for Railway Networks
1 Introduction
2 Data Set
3 COVID-19 Spreading Characteristics via Rail Network
3.1 Network Modeling
3.2 Basic Characteristics of the CHR Network
3.3 Spreading Characteristics Analysis
4 Conclusion
References
Learning to Collocate Fashion Items from Heterogeneous Network Using Structural and Textual Features
1 Introduction
2 Related Work
3 Fashion Collocation Based on Heterogenous Network
3.1 Overview of Our Framework
3.2 Network Construction
3.3 Structural Feature Extraction
3.4 Textual Feature Extraction
3.5 Feature Fusion
4 Experiment
4.1 Dataset
4.2 Experiment Settings
4.3 Results and Comparison
4.4 Parametric Study
5 Conclusion
References
Building Energy Performance Certificate Labelling Classification Based on Explainable Artificial Intelligence
1 Introduction
2 Problem Formulation
3 Methodology
3.1 ANN Modelling
3.2 Model Training, Test, and Evaluation
3.3 Explanation of the Building EPC Labelling Classification Model
3.4 Model Improvement and Optimisation
4 Case Study
4.1 Data Description and Processing
5 Results and Discussions
5.1 Trained Model Analysis
5.2 LIME XAI Results
6 Conclusion
References
Cross Languages One-Versus-All Speech Emotion Classifier
1 Introduction
2 Related Work
3 Methodology
3.1 Overall Structure
3.2 Feature Extraction
3.3 Feature Engineering
4 Experimental Results
4.1 Experimental Setup
4.2 Results of Feature Engineering
4.3 Framework Evaluation
5 Conclusion
References
A Hybrid Machine Learning Approach for Customer Loyalty Prediction
1 Introduction
2 Related Work
3 Research Method
3.1 K-Means Clustering
3.2 Classification Models for Prediction
3.3 Design of Two-Stage Model
4 Data
4.1 Dataset
4.2 Feature Selection and Engineering
4.3 Data Analysis Process
4.4 K-Means Clustering
4.5 Building Classification Models
4.6 Model Evaluation Techniques
5 Experimental Results and Discussions
5.1 Model Performance Review
5.2 Decision Tree Formulation
6 Conclusion and Future Work
References
LDA-Enhanced Federated Learning for Image Classification with Missing Modality
1 Introduction
2 Related Work
2.1 Federated Learning
2.2 Pattern Recognition with Missing Modality
3 Proposed Method
3.1 Framework
3.2 LDA-Based Features Aggregation
4 Experiments
4.1 Results on the MNIST Dataset
4.2 Results on the CIFAR-10 Dataset
5 Conclusion
References
A Data Enhancement Method for Gene Expression Profile Based on Improved WGAN-GP
1 Introduction
2 Preliminaries
2.1 Conditional Generative Adversarial Networks
2.2 Wasserstein Generative Adversarial Network Based on Gradient Penalty
3 The Proposed Method
3.1 Dataset Partition
3.2 Constraint Penalty Term
3.3 The Steps of the Proposed Method
4 Experiments and Discussion
4.1 Datasets and Algorithm Parameters Setting
4.2 Wasserstein Distance Index
4.3 Diversity Comparison on the Generated Sample with Different Methods
4.4 Stability Comparison on the Generated Sample Distribution Stability with Different Methods
4.5 Selection of the Threshold Parameter
5 Conclusions
References
Examining and Predicting Teacher Professional Development by Machine Learning Methods
1 Introduction
2 Related Works
3 The Proposed Questionnaire Scheme
4 Classification Problem and Machine Learning Methods
4.1 Classification Problem
4.2 Machine Learning Methods
4.3 Hyperparameter Optimization Scheme
5 Simulation Results
5.1 Identification of Significant Attributes
5.2 The Effect of Eight ML Methods
5.3 The Effect of Tuning ENS
5.4 The Effect of Tuning SVM
5.5 The Effect of Tuning ANN
5.6 Applying the ABC Algorithm to Tune ANN
6 Conclusion
References
Neural Computing-Based Fault Diagnosis, Fault Forecasting, Prognostic Management, and System Modeling
A Hybrid Approach to Risk Analysis for Critical Failures of Machinery Spaces on Unmanned Ships by Fuzzy AHP
1 Introduction
2 Preliminaries
2.1 Fuzzy Sets and Triangular Fuzzy Numbers
2.2 Z-numbers
3 Methodology
4 Case Study
4.1 Risk Measurement
4.2 Analysis of Risk of Black-Outs
5 Discussion and Recommendations
6 Conclusion
References
A New Health Indicator Construction Approach and Its Application in Remaining Useful Life Prediction of Bearings
1 Introduction
2 Theory Background
2.1 Spectral Clustering
2.2 Trendability
3 Health Indicators Construction Based on Spectral Clustering and Trendability Enhancement Strategy
3.1 Feature Extraction
3.2 Feature Clustering Based on Spectral Clustering
3.3 Sensitive Feature Selection Based on Trendability Optimization
3.4 Health Indicator Construction
4 Case Study
4.1 Data Description
4.2 Construction of HI-SCTWF
4.3 Remaining Useful Life Prediction
4.4 Comparison of Different Health Indicators
5 Conclusions
References
An Improved Reinforcement Learning for Security-Constrained Economic Dispatch of Battery Energy Storage in Microgrids
1 Introduction
2 Optimization Problem Modeling
2.1 Problem Description
2.2 Modeling for Reinforcement Learning
3 Proposed Algorithm
3.1 Policy Net with Domain-Knowledge-Based Rules
3.2 Distributional Critic Net
3.3 Protection Layer Control Method
4 Case Study
4.1 System Description
4.2 Numerical Results
5 Conclusion
References
Self-supervised Learning Advance Fault Diagnosis of Rotating Machinery
1 Introduction
2 Methods
2.1 Wavelet Transform for Data Pre-processing
2.2 Self-supervised Pretraining
3 Experiment Setup and Results
3.1 Self-supervised Pretraining
3.2 Supervised Fine-Tuning
3.3 Experimental Results
4 Conclusion
References
A Method for Imbalanced Fault Diagnosis Based on Self-attention Generative Adversarial Network
1 Introduction
2 Theoretical Background
2.1 Continuous Wavelet Transform
2.2 Generative Adversarial Networks
2.3 Self-attention Mechanism Module
2.4 Spectral Normalization
2.5 Frechet Inception Distance
3 System Framework
3.1 Data Preparation
3.2 The Architecture of the Generator and Discriminator
3.3 Fault Diagnosis Classifier
4 Case Study and Experiment Result
4.1 Introduction of Experimental Datasets
4.2 Images Generation
4.3 Fault Diagnosis Classification
5 Conclusion
References
Self-supervised Contrastive Representation Learning for Machinery Fault Diagnosis
1 Introduction
2 Method Definition
2.1 Contrastive Predicting Coding
2.2 Autoencoder
2.3 Statistical Features
2.4 Evaluation Metrics
3 Experimental Evaluation
3.1 Dataset
3.2 Model Training
3.3 Evaluation Using Dimensionality Reduction
3.4 Evaluation Using Clustering
3.5 Discussion of Results
4 Summary
References
SGWnet: An Interpretable Convolutional Neural Network for Mechanical Fault Intelligent Diagnosis
1 Introduction
2 Theoretical Basis
2.1 Second Generation Wavelet Transform
2.2 Estimation Method of Predictor and Updater Coefficients Based on Equivalent Filter
2.3 CNN
3 Methods of the Paper
3.1 Combination of Second Generation Wavelet and Convolutional Neural Network
3.2 Framework and Parameters of SGWnet
4 Simulation Experiment
4.1 Simulation Experiment Based on Inner Product Matching Principle
4.2 Validation of CWRU Dataset
5 Conclusion
References
GFU-Net: A Deep Learning Approach for Automatic Metal Crack Detection
1 Introduction
2 Related Work
2.1 Traditional Crack Detection Methods
2.2 Deep Learning-Based Methods
3 Methods
3.1 Overview of Proposed Method
3.2 Guide Transformer
3.3 Data Augmentation
4 Experiments
4.1 Experimental Setting
4.2 Ablation Study
4.3 Evaluations
5 Conclusion
References
Computational Intelligence, Nature-Inspired Optimizers, and Their Engineering Applications
An Improved Cluster Load Balancing Scheduling Algorithm
1 Introduction
2 Related Work
3 Algorithmic Implementation
3.1 Scheduling Algorithm Model and Overall Process
3.2 Measurement of Server Node Load
3.3 Resource Balance Modelling
3.4 Design of the Adaptation Function
4 Simulation
4.1 Test Environment
4.2 Response Delay
4.3 Throughput
4.4 Request Error Rate
4.5 Mean Variance in Resource Utilization
5 Summary
References
An Improved Cloud Particles Optimizer for Function Optimization
1 Introduction
2 Related Work
2.1 Cloud Gaseous Phase
2.2 Cloud Liquid Phase
2.3 Cloud Solid Phase
3 Improved Cloud Particles Optimizer
3.1 Fluid Operation
3.2 Solid Operation
3.3 Control Parameters Selection Mechanism
4 Experiments and Discussions
4.1 General Experimental Setting
4.2 Comparison of ICPEA with Other Optimization Algorithms
5 Conclusion
References
Resource Allocation and Trajectory Optimization for UAV Assisted Mobile Edge Computing Systems with Energy Harvesting
1 Introduction
2 System Model and Problem Formulation
2.1 UAV's Trajectory Model
2.2 Communication Channel Model
2.3 Energy Consumption and MUs' Task-Input Model
2.4 Problem Formulation
3 Problem Transformation and Solution
3.1 Optimization of the UAV's Transmission Power and Time Allocation
3.2 Optimization of the Trajectory of the UAV
3.3 Overall Algorithm
4 Numerical Simulation Results
4.1 Trajectory of the UAV
4.2 Performance Comparison
5 Conclusion
References
Meta-feature Extraction for Multi-objective Optimization Problems
1 Introduction
2 Related Background
2.1 Meta-feature
2.2 Pareto Front Geometrical Features of MOPs
3 Proposed Meta-features for MOPs
3.1 Target Space-Based Features
3.2 PF-Based Features
4 Experimental Validation and Results
4.1 Experimental Process
4.2 Experimental Setup
4.3 Experimental Results and Evaluation
5 Conclusion
References
Feed Formula Optimization Based on Improved Tabu Search Algorithm
1 Introduction
2 Related Work
2.1 Feed Formula Optimization Problem
2.2 Tabu Search Algorithm
3 Improved Tabu Search Algorithm for Feed Formula Optimization Problem
3.1 Feed Formula Optimization Problem
3.2 Improved Tabu Search Algorithm
4 Experiments
4.1 Comparison Between the ITS Algorithm with Intelligent Optimization Algorithms
4.2 Comparison Between Algorithms Using Innovative Feed Formula
4.3 Analysis of the ITS Algorithm
4.4 Analysis of the Experiment Results
5 Conclusions
References
Adaptive Methods of Differential Evolution Multi-objective Optimization Algorithm Based on Decomposition
1 Introduction
2 Background
2.1 Basic Definitions
2.2 MOEA/D
2.3 Differential Evolutionary Algorithms
3 Proposed Algorithms
3.1 DE in MOEA/D
3.2 Adaptive Operators
3.3 The Framework of the Proposed Algorithms
4 Experimental Research and Results Analysis
4.1 Benchmark Problems
4.2 Parameter Settings
4.3 Performance Metrics
4.4 Experiments Analysis
5 Conclusion
References
Population Diversity Guided Dimension Perturbation for Artificial Bee Colony Algorithm
1 Introduction
2 Artificial Bee Colony Algorithm
3 Proposed Approach
4 Experimental Study
4.1 Experimental Design and Parameter Selection
4.2 Results on Classic Benchmark Problems
4.3 Results on CEC 2013 Benchmark Problems
5 Conclusion
References
Artificial Bee Colony Algorithm with an Adaptive Search Manner
1 Introduction
2 The Original Artificial Bee Colony Algorithm
3 ABC with Adaptive Search Manner (ASMABC)
3.1 Strategy Pool
3.2 Adaptive Search Manner
4 Experimental Study
5 Conclusion
References
Fuzzy Logic, Neuro-Fuzzy Systems, Decision Making, and Their Applications in Management Sciences
Active Learning Method Based on Axiomatic Fuzzy Sets and Cost-Sensitive Classification
1 Introduction
2 Preliminaries
2.1 AFS Algebra
2.2 AFS Membership Function
2.3 Neighbor and Mutual Nearest Neighbor
2.4 Cost-Sensitive Decision System
2.5 Misclassification Cost Matrix
2.6 Select Critical Instances
3 The Designed Method
4 Experiments
4.1 Case Study
4.2 Experiment Results
5 Conclusions
References
An Extended TODIM Method Based on Interval-Valued Pythagorean Hesitant Fuzzy Sets and Its Utilization in Green Shipping
1 Introduction
2 Preliminaries
2.1 Interval-Valued Pythagorean Hesitant Fuzzy Sets
2.2 The Classical TODIM Method
3 Cosine Similarity Measure of Interval-Valued Pythagorean Hesitant Fuzzy Sets
4 Solving Multi-attribute Group Decision-Making Problem Based on Extended TODIM Method
4.1 The Extended TODIM Method
4.2 Operation Steps of the Designed Method
5 Numerical Study
5.1 Ranking Evaluation Results of Container Ships
5.2 Experimental Verification
5.3 Sensitivity Analysis
6 Conclusion
References
Control Systems, Network Synchronization, System Integration, and Industrial Artificial Intelligence
Linear Time-Varying Model Predictive Control for Trajectory-Tracking of a Wheeled Mobile Robot
1 Introduction
2 Problem Formulation
2.1 Kinematic Model of the Wheeled Mobile Robot
2.2 Control Objective
3 Linear Time-Varying MPC for the Wheeled Mobile Robot
3.1 Linearization and Discretization
3.2 Linear Time-Varying MPC Design
4 Stability Analysis on the Closed-Loop System
5 Simulation
6 Conclusion
References
A Traffic Light Control System Based on Reinforcement Learning and Adaptive Timing
1 Introduction
2 Related Work
3 Method
3.1 Proposed Traffic Light Control System
3.2 Agent Design
3.3 A2C Algorithm
3.4 Adaptive Timing Algorithm
4 Experiment
4.1 Experiment Setting and Parameter Setting
4.2 Evaluation Metric
4.3 Compared Methods
4.4 Performance on Synthetic Data
4.5 Performance of a Whole Day
5 Conclusion
References
Computer Vision, Image Processing, and Their Industrial Applications
A Transformer-Based Decoupled Attention Network for Text Recognition in Shopping Receipt Images
1 Introduction
2 Related Work
2.1 Text Detection
2.2 Text Recognition
3 Our Method
3.1 Overview of Our Framework
3.2 Multi-task Text Detection Model
3.3 Transformer-Based Decoupled Attention Network
4 Experiment
4.1 Experiments on Multi-task Text Detection Model
4.2 Experiment on Transformer-Based Decoupled Attention Network
5 Conclusions
References
Interactive Clothes Image Retrieval via Multi-modal Feature Fusion of Image Representation and Natural Language Feedback
1 Introduction
2 Proposed Method
2.1 Semantic Extraction Module
2.2 Image Extraction Module
2.3 Fusion Module
2.4 Matching Objective Module
3 Simulations
3.1 Dataset
3.2 Experimental Parameters
3.3 Results
4 Conclusion
References
An Ultra-High Speed Gesture Recognition Algorithm Based on MobileNetV2
1 Introduction
2 Related Work
3 The Proposed Method
3.1 Improved MobileNetV2
3.2 Cross-Validation Training
3.3 Pruning of the Network Model
3.4 Fusion of Model Operators
3.5 Model Quantification
4 Experiment and Analysis
5 Conclusion
References
An Hybrid Model CMR-Color of Automatic Color Matching Prediction for Textiles Dyeing and Printing
1 Introduction
2 Related Work
3 CMR-Color Fusion Model
3.1 Feature Extraction
3.2 Feature Extraction from Dye Composition Spectral Data
3.3 Feature Fusion and Dyeing Recipe Prediction
4 Evaluation and Results
4.1 Datasets
4.2 Evaluation Metrics
4.3 Settings
4.4 Results
5 Conclusions
References
Obstacle Avoidance Algorithm for Mobile Robot Based on ROS and Machine Vision
1 Introduction
2 Visual Detection Algorithm
2.1 Improved Object Detection Algorithm Based on YOLO-v4
2.2 Loss Function
3 Obstacle Avoidance Algorithm for Navigation
3.1 Global Path Planning Algorithm
3.2 Obstacle Avoidance Algorithm of Local Path
4 Experimental Results and Analysis
5 Conclusion
References
A New Total Variation Denoising Algorithm for Piecewise Constant Signals Based on Non-convex Penalty
1 Introduction
2 Problem Statement
3 Denoising Algorithm
3.1 Design of the Non-convex Penalty Function
3.2 Denoising Algorithm Design and Convergence Analysis
4 Experimental Results
5 Conclusion
References
Superpixel Segmentation via Contour Optimized Non-Iterative Clustering
1 Introduction
2 Related Work
2.1 Seed-Demand Superpixel Segmentation
2.2 Graph-Based Superpixel Segmentation
3 Contour Optimized Non-Iterative Clustering
3.1 Preliminaries on Non-Iterative Clustering
3.2 Improved Cluster Distance Measurement
4 Experimental Results and Analysis
4.1 Visual Assessment
4.2 Metric Evaluation
4.3 Running Efficiency
5 Conclusions
References
Semantic Segmentation via Efficient Attention Augmented Convolutional Networks
1 Introduction
2 Related Work
2.1 Convolutional Networks for Semantic Segmentation
2.2 Works to Decrease the Quadratic Complexities of Self Attention
3 Method
3.1 Efficient Attention
3.2 Column-Row Attention
3.3 Efficient Attention Augmented Convolution Module
4 Experiments
4.1 Comparison with FCN
4.2 Ablation Study
5 Conclusion
References
Cervical Spondylotic Myelopathy Segmentation Using Shape-Aware U-net
1 Introduction
2 Relate Work
3 Method
3.1 Shape-Aware Module
3.2 SAPP Module
4 Evaluation and Results
4.1 Implementation Details
4.2 Ablation Analysis of SAM and SAAP Module
4.3 Comparison to the Competing Methods
5 Conclusion
References
Cloud/Edge/Fog Computing, The Internet of Things/Vehicles (IoT/IoV), and Their System Optimization
An Efficient CSI-Based Pedestrian Monitoring Approach via Single Pair of WiFi Transceivers
1 Introduction
2 Related Work
3 System Architecture
4 Proposed Algorithm
4.1 CSI Pre-processing
4.2 Two-Stage Clustering Method
4.3 Pedestrian Pass Detection Algorithm
4.4 Pedestrian Pass Detection Algorithm
5 Performance Evaluation
5.1 Metrics
5.2 System Deployment and Parameter Setting
5.3 Adaptive Evaluation
6 Conclusion
References
Optimal Path Planning for Unmanned Vehicles Using Improved Ant Colony Optimization Algorithm
1 Introduction
2 Path Planning Using ACO Algorithm
2.1 Classical Ant Colony Optimization Algorithm
2.2 Application of Basic Ant Colony Optimization Algorithms
3 Path Planning Using Improved ACO Algorithm
3.1 Task Environment Modeling
3.2 Improvement of Pheromone Volatilization Coefficient
3.3 The Flow of Improved Ant Colony Algorithm
4 The Experimental Results
5 Conclusions
References
Spreading Dynamics, Forecasting, and Other Intelligent Techniques Against Coronavirus Disease (COVID-19)
Daily PM2.5 Forecasting Using Graph Convolutional Networks Based on Human Migration
1 Introduction
2 Datasets and Study Area
2.1 Air Pollution Data
2.2 AutoNavi Migration Data
2.3 Study Area
3 Methodologies
3.1 Graph Representing
3.2 Graph Convolutional Network
3.3 Graph Attention Network
4 Experimental Result
4.1 Experimental Settings
4.2 Evolution Criteria
4.3 Forecasting Results
5 Conclusion
References
A Novel Approach to Ship Operational Risk Analysis Based on D-S Evidence Theory
1 Introduction
2 Methodology
2.1 Entropy Weight Method
2.2 Intuitionistic Fuzzy Set
2.3 D-S Evidence Theory
3 Case Study
4 Conclusion
References
Short-Term Building Load Forecast Based on Patch Learning with Long Short-Term Memory Network and Support Vector Regression
1 Introduction
1.1 Background and Motivation
1.2 Literature Review
1.3 Novelty and Contributions
2 PL-LSTM-SVR Based Load Forecasting
2.1 STLF Problem Formulation
2.2 Patch Learning Framework
2.3 Global Model
2.4 Identifying Patch Locations
2.5 Patch Model
3 Case Study
3.1 Data Description
3.2 Setup
3.3 Result and Discussions
4 Conclusion
References
Empirical Mode Decomposition Based Deep Neural Networks for AQI Forecasting
1 Introduction
2 Methodologies
2.1 Empirical Mode Decomposition
2.2 1D Convolutional Neural Network
2.3 The Proposed EMD-Hybrid Model
3 Experimental Results
3.1 Dataset and Preprocessing
3.2 Evolution Criteria
3.3 Comparative Models
3.4 Forecasting Results
4 Conclusion
References
Author Index