Advances in Artificial Intelligence and Security: 8th International Conference on Artificial Intelligence and Security, ICAIS 2022, Qinghai, China, ... in Computer and Information Science, 1586) Part I

Preface
Organization
Contents – Part I
Contents – Part II
Contents – Part III
Artificial Intelligence
A Target Extraction Algorithm Based on Polarization Image Attention Mechanism
	1 Introduction
	2 Algorithm
		2.1 Feature Extraction Network
		2.2 Attention Mechanism
		2.3 Semi Supervised Transfer Learning
	3 Experiment and Result Analysis
		3.1 Image Information Statistics
		3.2 Target Extraction Result Statistics
		3.3 Network Performance of Transfer Learning
		3.4 Effect Display of Target Extraction
	4 Conclusion
	References
Fidelity-Preserved Reversible Data Hiding in JPEG Images Based on Adaptive Position Selection
	1 Introduction
	2 Previous Arts
		2.1 Overview of JPEG Compression Standard
		2.2 Overview of Xuan et al. Method
	3 The Proposed Method
		3.1 Minimum Entropy-Based Block Selection and Block Order
		3.2 AC Coefficient Position Selection and Position Order
		3.3 Embedding, Extraction and Restoration
	4 Experiment Results and Analyses
		4.1 Effectiveness of the Proposed Method
		4.2 Comparisons
	5 Conclusion
	References
Research on Image Multi-feature Extraction of Ore Belt and Real-Time Monitoring of the Tabling by Sema ntic Segmentation of DeepLab V3 +
	1 Introduction
	2 Materials and Method
		2.1 Acquisition and Production of Data Sets
		2.2 Construction of Semantic Segmentation Model by Deep Learning
		2.3 Train and Prediction of DeepLab V3 + model
		2.4 Evaluation of the Results
	3 Real-Time Monitoring and Intelligentization of Tabling
		3.1 Modeling the Relationship Between Image Features and Indexes of Tabling Processing
		3.2 Conception of Real-Time Monitoring and Intelligent System
	4 Conclusions
	References
A Reinforcement Learning-Based Method to Coordinated Multi-energy Optimization
	1 Introduction
	2 Related Work
	3 Multi-energy Synergy Optimization
		3.1 Optimization Objectives
		3.2 Constraints
	4 Collaborative Optimization of Multiple Energy Sources Based on RL
		4.1 Neural Network Structure
	5 Experimental Results and Analysis
	6 Conclusion
	References
Real-Time Domestic Garbage Detection Method Based on Improved YOLOv5
	1 Introduction
		1.1 Related Work
		1.2 Main Work
	2 YOLOv5 Algorithm
	3 Improvement of YOLOv5 Network Model
		3.1 Diou-Nms
		3.2 ACON-C Activation Function
	4 Experiments
		4.1 Data Set
		4.2 Experiment Environment
		4.3 Experiment Environment
	5 Conclusion
	References
Research on Epidemic Spreading Model Based on Double Groups
	1 Introduction
	2 Current Research Status at Home and Abroad
		2.1 Propagation Model Based on Statistical Analysis
		2.2 Propagation Model Based on Statistical Analysis
	3 Model Definition and Stability Analysis
	4 Experiment
		4.1 The Impact of the Incubation Rate 1/2on the Spread of the Epidemic
		4.2 The Impact of 1/2Recovery Rate on the Spread of the Epidemic
	5 Summary
	References
Application Prospect of High-Precision Map Mode and Pure Vision Mode in Automatic Driving
	1 Introduction
	2 The Significance of High-Precision Maps in Autonomous Driving
	3 Application and Limitation Analysis of High-Precision Maps in Autonomous Driving
	4 Autonomous Driving Solution Based on Pure Vision
	5 Conclusion
	References
An Automatic Construction Technology of Chinese Knowledge Graph for Teaching Plan
	1 Introduction
	2 Research Status of Knowledge Graph in the Field of Education
		2.1 Knowledge Annotation
		2.2 Knowledge Extraction
		2.3 Entity Alignment
		2.4 Graph Generation
	3 Automatic Construction of Knowledge Graph of Teaching Plan
		3.1 Raising Questions
		3.2 Implementation Process
	4 Implementation
		4.1 Extracting Knowledge Point Entities
		4.2 Extract the Relationship Between Knowledge Points and Entities
		4.3 Relationship Extraction Rules
	5 Testing and Summary
	References
Research on Cloud Office Resource Allocation Algorithm Based on Correction Weight PSO
	1 Introduction
	2 Related Work
		2.1 Cloud Office Services
		2.2 Particle Swarm Optimization
	3 Cloud Office Resource Optimization Algorithm
		3.1 Cloud Office Service Model
		3.2 Correction Weight PSO (CWPSO)
	4 Simulation Verification
	5 Conclusion
	References
Analysis of Wind Speed Characteristics During Typhoon Rammasun: A Case Study of Qinzhou, Guangxi
	1 Introduction
	2 Research Methods
		2.1 Gust Factor
		2.2 Wind Shear Index
	3 Analysis of Wind Speed Characteristics of Typhoon Rammasun
		3.1 Analysis of Gust Coefficient of Wind Tower
		3.2 Analysis of Wind Shear Index Based on Lidar
		3.3 Air Density Correction
	4 Conclusion
	References
An MR Image Segmentation Method Based on Dictionary Learning Preprocessing and Probability Statistics
	1 Introduction
	2 Method
		2.1 Bounding Box
		2.2 Data Labelling
		2.3 Bayesian Model Training
		2.4 Dictionary Leaning Denoising
		2.5 Segmentation
	3 Experimental Results
	4 Conclusion
	References
D2D Resource Allocation Problem in Full - Load Cellular Networks
	1 Introduction
		1.1 Related Work
		1.2 Contribution
		1.3 Organization of the Paper
	2 System Model and Problem Description
		2.1 System Model
		2.2 Power Distribution and Path Loss
		2.3 Problem Description
	3 Design of the D2D Link Sharing Algorithm Based on Throughput and Interference Reduction
		3.1 Link Candidates
		3.2 Further Interference Screening
	4 Simulation Results and Analysis
		4.1 Simulation Parameters and Simulation Process
		4.2 Influence of Channel Fading on D2D Communication Performance
	5 Conclusion
	References
Study on the Transmission Characteristics and the Longest Transmission Distance of 0.1–1 Terahertz Wave in the Atmosphere
	1 Introduction
	2 Transmission Characteristics of Terahertz Wave in Atmosphere
		2.1 Absorption Model of Oxygen and Water Vapor and Solution
		2.2 Non Resonant Continuous Absorption Model of Dry Air and Water Vapor and Solution
	3 Simulation of Terahertz Wave Transmission Distance
	4 Conclusion
	References
Research on Diabetic Retinal Fundus Screening Model Based on Conditional Generative Adversarial Nets
	1 Introduction
	2 Problem Description
		2.1 Model Structure
		2.2 Generate Network
		2.3 Discriminant Network
	3 Solution Method
		3.1 Discriminant Network
		3.2 Optimization of Model Training Gradient
		3.3 Correction and Solution of Exudate Prediction Results
	4 Experiment and Analysis
		4.1 Data Preparation and Operation Environment
		4.2 Evaluation Criteria
		4.3 Analysis of Results
	5 Conclusion
	References
One-Hot Coding Similarity of Vehicle Routing Particle Swarm Optimizer
	1 Introduction
	2 Vehicle Routing Problem
		2.1 Vehicle Routing Problem
		2.2 Components of VRP
		2.3 Mathematical Model of the Capacity Vehicle Routing Problem
	3 PSO for Vehicle Routing Problem
		3.1 Particle Swarm Optimization (PSO)
		3.2 The Encoding Methods
		3.3 Constraint Processing
		3.4 Particle Swarm Optimization Algorithm for Adaptive Vehicle Routing Particle
	4 Experimental Results and Analysis
		4.1 Case 1
		4.2 Case 2
	5 Conclusion
	References
Detection Method of Potholes on Highway Pavement Based on Yolov5
	1 Introduction
	2 Related Work
	3 Method
		3.1 Dataset
		3.2 Network Structure
		3.3 Evaluation Metrics
	4 Experiment
		4.1 Experimental Details
		4.2 Experimental Results
	5 Conclusion
	References
Further Understanding Towards Sparsity Adversarial Attacks
	1 Introduction
	2 Related Work
		2.1 Attacking Methods
		2.2 Defense Methods
	3 Proposed Methods
		3.1 Background
		3.2 BPs Method
		3.3 Binary Fitting Method
	4 Experiments
		4.1 Settings
		4.2 Evaluation of Targeted Attack Ability
		4.3 Evaluation of Untargeted Attack Ability
	5 Conclusion
	References
MRDA-Net: Multiscale Residual Dense Attention Network for Image Denoising
	1 Introduction
	2 Related Work
		2.1 Dilated Convolution
		2.2 Dense Connection
		2.3 Attention Mechanism
	3 Proposed Method
		3.1 Network Structure
		3.2 Implementation
	4 Experiment
		4.1 Datasets
		4.2 Ablation Studies
		4.3 Non-blind Image Denoising
		4.4 Blind Image Denoising
		4.5 Complexity of Network
	5 Conclusion
	References
Aspect-Level Sentiment Classification Based on Graph Attention Network with BERT
	1 Introduction
	2 Related Works
	3 The Proposed Method
		3.1 Task Definition
		3.2 The Whole Architecture
		3.3 Attention Module
		3.4 Graph Attention Network
		3.5 Sentiment Classification Layer
	4 Experiment and Results
		4.1 Dataset
		4.2 Training and Parameters Setting
		4.3 Models for Comparison
		4.4 Results and Analysis
	5 Conclusion
	References
Design and Implementation of Teacher Workload Management System Based on JQuery Mobile
	1 Introduction
	2 Technical Selection and Requirement Analysis
		2.1 Framework to Choose
		2.2 System Modeling
		2.3 System Function Module Design
	3 Implementation Process and Results
		3.1 Workload Management
		3.2 Overtime Leave Management Module
		3.3 Workload Retrieval and Modification Module
		3.4 Graduation Design Workload Statistics Module
		3.5 Workload Audit Module
		3.6 Project Management Module
		3.7 System Key Technology
		3.8 The System Test
	4 Usage and Summary
		4.1 Use Effect
		4.2 Conclusion
	References
A PSAO-Based Optimization Method for Integrated Energy Systems in Cities and Towns
	1 Introduction
		1.1 A Subsection Sample
	2 Related Work
	3 Regional Integrated Energy System Planning and Operation Optimization
		3.1 Objective Function
		3.2 Constraints
		3.3 Methods
	4 Experimental Results and Analysis
		4.1 Summer Optimization Results
		4.2 Winter Optimization Results
		4.3 Convergence Verification
	5 Conclusion
	References
Applied Research on Cultivating Advanced Technical Engineering Talents Based on the Internet of Things+ Platform with Dual-Professional Teachers as the Guidance
	1 Introduction
	2 Review of Research at Home and Abroad
	3 The Mode of Training Advanced Technical Talents
		3.1 Internet of Things+ Platform
		3.2 The Guidance of Double Qualified Teachers
		3.3 Cultivation of Higher Technical Talents
	4 Conclusion
	References
Classification of Tropical Cyclone Intensity Based on Deep Learning and YOLO V5
	1 Introduction
	2 Method
		2.1 The Input
		2.2 YOLO V5 Backbone Network
		2.3 Loss Function
	3 Experiment and Result Analysis
		3.1 Experimental Dataset
		3.2 Experimental Setup
		3.3 Algorithm Evaluation Indicators
	4 Experimental Results
		4.1 Comparison and Analysis of Transfer Learning Results
	5 Analysis and Discussion
		5.1 In a Continuous Sequence, Tropical Cyclones in the Middle Cannot Be Identified
		5.2 Some Tropical Cyclones Identify Two Possibilities
	6 Conclusion
	References
STKE: Temporal Knowledge Graph Embedding in the Spherical Coordinate System
	1 Introduction
	2 Related Work
	3 Proposed Model
		3.1 STKE
		3.2 Loss Function
	4 Experiments and Analysis
		4.1 Datasets
		4.2 Experimental Settings
		4.3 Experimental Results
		4.4 Analysis
	5 Conclusion
	References
Improved VGG-16 Neural Network for Parameter Reduction
	1 Introduction
	2 Related Work
		2.1 Convolutional Neural Networks
		2.2 Depthwise Separable Convolutions
	3 VGG-16-JS
	4 Experiments and Analysis
	5 Conclusion
	References
Directly Recognize Who a Specific Pedestrian is with a Surveillance Camera
	1 Introduction
	2 Method
		2.1 Specific Pedestrian Recognition System
		2.2 Kernelized Correlation Filters
		2.3 Yolov4-Based Face Detection Module
		2.4 Face Recognition Based on ResNet-50 and ArcFace
	3 Experiment
		3.1 Face Recognition Based on ResNet-50 and ArcFace
		3.2 Comparison Between Anchor-Based Yolov4 and Anchor-Free Yolov4
	4 Conclusion
	References
Minimizing Immune Costs in Social Networks Through Reinforcement Learning
	1 Introduction
	2 Related Work
	3 Related Work
		3.1 The Network Model
		3.2 Propagation Model
		3.3 Reinforcement Learning Networks
	4 Immune Cost Algorithm
		4.1 Problem Define
		4.2 Propagation Model
		4.3 Case Analysis
	5 Simulation Study
		5.1 Simulation Configure
		5.2 Simulation Results
	6 Conclusion
	References
End-to-End Speech Synthesis Method for Lhasa-Tibetan Multi-speaker
	1 Introduction
	2 Method
		2.1 Text Preprocessing
		2.2 Spectrogram Feature Prediction Network
		2.3 Vocoder
	3 Experiments Settings
	4 Experimental Results
		4.1 Subjective Experiment
		4.2 Objective Experiment
	5 Summary
	References
Prediction Model of Optimized Grey Neural Network Based on Mind Evolutionary Algorithm
	1 Introduction
	2 Mind Evolutionary Algorithm
	3 MEA Optimized Grey Neural Network
		3.1 Grey Neural Network
		3.2 MEA Optimization Grey Neural Network Algorithm
	4 Design and Analysis of Experimental Results
		4.1 Experimental Design
		4.2 Analysis of Experimental Results
	5 Conclusion
	References
An End-to-End Multi-dialect Tibetan Speech Synthesis Method
	1 Introduction
	2 Model Structure
		2.1 Tibetan Script
		2.2 Encoder
		2.3 Decoder
		2.4 Attention Mechanism
		2.5 WaveNet Vocoder
	3 Experiments and Results
		3.1 Experimental Data
		3.2 Comparative Experiments
		3.3 Results
	4 Conclusions
	References
A Hybrid Automatic Text Summarization Model for Judgment Documents
	1 Introduction
	2 Related Work
	3 Method
		3.1 Extractive Model
		3.2 Abstractive Model
	4 Experiments
		4.1 Datasets
		4.2 Implementation Details
		4.3 Experimental Results and Analysis
	5 Conclusion
	References
Quantum Fuzzy Principal Component Analysis
	1 Introduction
	2 Principal Component Analysis
	3 Quantum Fuzzy Principal Component Analysis
		3.1 Constructing Fuzzy Covariance Matrix
		3.2 Constructing Fuzzy Covariance Matrix
		3.3 Quantum Circuit Implementation
	4 Analysis
		4.1 Runtime Analysis
		4.2 Effectiveness Analysis
	5 Conclusions
	References
A Handwritten Number Recognition Scheme Based on Improved Convolutional Neural Network Algorithm
	1 Introduction
	2 Related Work
	3 Implementation and Problems of Key Algorithms
		3.1 Implementation of K Nearest Neighbor Algorithm
		3.2 Implementation of Convolutional Neural Network Algorithm
		3.3 Differences of the Two Algorithms
	4 Optimization and Improvement of Key Algorithms
		4.1 The Improvement of K Nearest Neighbor Algorithm and Its Effect
		4.2 Improvement and Effect of Convolutional Neural Network Algorithm
	5 Handwritten Digit Recognition System Based on Machine Learning
		5.1 Handwritten Digit Recognition Development Environment
		5.2 Handwritten Digit Recognition Development Environment
		5.3 Handwritten Digit Recognition Development Environment
		5.4 Handwritten Digit Recognition Development Environment
	6 Concluding Remarks
	References
D3: A Novel Face Forgery Detector Based on Dual-Stream and Dual-Utilization Methods
	1 Introduction
	2 Related Works
		2.1 Face Forgery
		2.2 Face Forgery Detection
	3 Methods
		3.1 Temporal Stream
		3.2 Spatial Stream
	4 Experiment
		4.1 Datasets
		4.2 Implementation Details
		4.3 Comparison Experiment
		4.4 Ablation Experiment
		4.5 Cross-dataset Experiment
	5 Conclusion
	References
Surface Defect Detection of Medium and Thick Plates Based on MASK-RCNN
	1 Introduction
		1.1 A Subsection Sample
	2 Algorithmic Principles
		2.1 BACKBON
		2.2 RPN Network
		2.3 ROI
		2.4 Loss
	3 Methodology
		3.1 Image Acquisition
		3.2 Image Overlays
		3.3 Data Enhancement
		3.4 Experimental Environment
		3.5 Parameter Setting
		3.6 Evaluating Indicator
	4 Results
	5 Conclusion
	References
Video-Based Pedestrian Re-identification with Non-local Attention Module
	1 Introduction
	2 Related Work
		2.1 Attentional Mechanism
	3 Video-Based Person Re-identification with Non-local Attention Module
		3.1 Non-local Mean Module (NLMM)
		3.2 Self-attention Module (SAM)
		3.3 Non-local Attention Module (NLAM)
	4 Experiments and Results
		4.1 Datasets
		4.2 Experimental Parameter Settings
		4.3 Experimental Evaluation
	5 Summary
	References
Long-Tailed Classification Based on Dual Branch Learning Network
	1 Introduction
	2 Overall Framework
		2.1 Long-Tail Recognition Model
		2.2 Re-weighting Branch
		2.3 Data Augmentation Learning Branch
	3 Experiment
		3.1 Datasets
		3.2 Baseline Methods
		3.3 Implementation Details
		3.4 Results on Imbalanced CIFAR
		3.5 Results on Rail Fastener Dataset
	4 Conclusion
	References
Hamming Code Aided Joint Iterative Detection and Decoding Receiver of Polar Coded SCMA System
	1 Introduction
	2 System Model
	3 Hamming Code Aided Joint Iterative Detection and Decoding Algorithm
		3.1 Scheme Description
		3.2 Complexity Analysis
	4 Results and Discussion
	5 Conclusion
	References
Recognition of Pedestrians’ Crossing Intentions with a Conv-Transformer Network
	1 Introduction
	2 The Proposed Method
		2.1 Spatial Features Extraction
		2.2 Temporal Features Extraction
		2.3 Loss Function
	3 Experimental Evaluations
		3.1 Databases
		3.2 Evaluation Metric
		3.3 Experimental Details
		3.4 Quantitative Evaluations
		3.5 Qualitative Evaluations
	4 Conclusion and Discussion
	References
Violent Target Detection Based on Improved YOLO Network
	1 Introduction
	2 Related Work
	3 Models and Methods for Detection of Riot-Related Images
		3.1 YOLOv5x
		3.2 Yolov5 Fusion Attention Mechanism
		3.3 Loss Function Using CIOUloss
	4 Experimental Procedure and Results
		4.1 Experimental Preparation
		4.2 Experimental Results and Comparative Analysis
	5 Conclusion
		5.1 Online Violence-Related Content Remains a Concern and a Solution
		5.2 Existing Problems and Future Prospects
		5.3 Application Implications
	References
Analysis and Architecture Design of a Large-Scale Event-Centric Knowledge Graph System for Dispute Resolution
	1 Introduction
	2 System Architecture Design
	3 Data Preprocessing
	4 Methods of Building KG
		4.1 Ontology Building Module
		4.2 Event Extraction Module
		4.3 Knowledge Graph Refinement Module
	5 Conclusions
	References
Quantum Fuzzy Least Squares Algorithm for Uncertain Environment
	1 Introduction
	2 Related Work
		2.1 Least Square Method
		2.2 Quantum Least Squares Method
		2.3 Fuzzy Least Squares Method
	3 Quantum Fuzzy Least Squares Algorithm
	4 Analysis of Algorithms
		4.1 Time Complexity Analysis
		4.2 Accuracy Analysis
	5 Concluding Remarks
	References
Research on Instruction Pipeline Optimization Oriented to RISC-V Vector Instruction Set
	1 Introduction
	2 Background and Related Work
	3 Instruction Pipeline Optimization Method
	4 Performance Evaluation
	5 Citations
	References
The Nexus Between Air Pollution and Life Insurance Demand in China: Evidence from Deep Machine Learning
	1 Introduction
	2 Literature Review
		2.1 The Influencing Factors of Life Insurance Demand
		2.2 The Influence of Air Pollution on Insurance Purchase Decision
	3 Model Assumption and Variable Selection
		3.1 Dependent Variable
		3.2 Independent Variables
		3.3 Control Variables
	4 Empirical Results
		4.1 Description of Variables and Data Sources
		4.2 Model Setting and Testing
		4.3 Empirical Results
		4.4 Further Demonstration Based on Deep Learning Method
	5 Explanation of Results
	6 Conclusion and Discussion
	References
Image Steganalysis Based on CNN and Transfer Learning for Small-Sample Condition
	1 Introduction
	2 Related Works
		2.1 Steganalysis Technology Based on Traditional Methods
		2.2 Steganalysis Technology Based on Deep Learning
	3 Method
		3.1 Problem Description
		3.2 Transfer Learning in Steganalysis
		3.3 Training Strategy
	4 Experiments
		4.1 Datasets
		4.2 Experiments Settings
		4.3 Results
	5 Conclusion
	References
Find the Unseen Actions: Abnormal Action Recognition
	1 Introduction
	2 Problem Description and Evaluation
	3 Method Introduction
		3.1 Softmax Distribution
		3.2 Mahalanobis Distance-Based Confidence Score
		3.3 Classification Feature
	4 Experiment
		4.1 Datasets
		4.2 Setups
		4.3 Results
	5 Discussion and Future Work
	6 Conclusion
	References
CGFMD: CNN and GRU Based Framework for Malicious Domain Name Detection
	1 Introduction
	2 Related Works
		2.1 Malicious Domain Name Detection
		2.2 Gate Recurrent Unit
		2.3 Natural Language Processing
	3 Method
		3.1 Data Encode
		3.2 Word Vector Mapping
		3.3 Algorithm Introduction
	4 Experiment
		4.1 Experimental Environment
		4.2 Experimental Setup
		4.3 Baselines
		4.4 Experimental Results
	5 Conclusion
	References
Quintuple Implication Principle on Intuitionistic Fuzzy Sets
	1 Introduction
	2 Preliminary Knowledge
	3 The Quintuple I Algorithm Based on Intuitionistic Fuzzy Sets
	4 The α-Quintuple I Algorithm Based on Based on Intuitionistic Fuzzy Sets
	References
A Module Based Full Cycle Construction Method of Domain-Specific Knowledge Graph
	1 Introduction
	2 Background and Related Work
		2.1 Ontology Development
		2.2 Information Extraction
		2.3 Entity Resolution
	3 Proposed Methods
		3.1 Architecture Overview
		3.2 Modules in Detail
	4 Case Study
	5 Conclusions
	References
A Transferable Framework for Few-Shot Human Activity Recognition
	1 Introduction
	2 Related Work
	3 Model and Methods
		3.1 Overall Framework
		3.2 Models
	4 Experiments Process and Results
		4.1 Benchmark Datasets
		4.2 Model Training
		4.3 Result
	5 Conclusion
	References
Incorporating Syntactic and Semantic Information into Compound Sentence Discrimination and Relation Recognition
	1 Introduction
	2 Related Work
	3 Model
		3.1 Input Representation
		3.2 Bi-LSTM
		3.3 Tree-LSTM
		3.4 Compound Sentence Discrimination
		3.5 Relation Recognition
		3.6 Joint Learning
	4 Experimentation
		4.1 Data Set
		4.2 Experiment Settings
		4.3 Experimental Results and Analysis
		4.4 Model Comparison and Analysis
	5 Conclusion
	References
Research on Side-Channel Attack Method Based on LSTM
	1 Introduction
	2 Related Work
		2.1 Advanced Encryption Standard
		2.2 Energy Analysis Attack
		2.3 Long Short-Term Memory Networks
	3 LSTM - Based Side Channel Attack
		3.1 Select the Data TAB
		3.2 Point of Interest Choose a Choice
	4 Experiment and Discussion
		4.1 Prepare the Laboratory Environment
		4.2 According to the Data Set
		4.3 Attack Effect
	5 Conclusion
	References
A Novel EEG-Based Depression Detection Framework
	1 Introduction
	2 Methods
		2.1 Data Acquisition
		2.2 Data Preprocessing
		2.3 Feature Selection
		2.4 Classification Algorithms
	3 Results and Discussion
		3.1 Computing Environment
		3.2 Classification Performance
	4 Conclusion
	References
Tunnel Disease Detection Based on Spark and Deep Learning
	1 Introduction
	2 Related Work
		2.1 Mass Data Processing Technology Based on Spark
		2.2 Approaches of Metro Disease Detection
	3 Distributed Parallel Point Cloud Data Processing Based on Spark
		3.1 Space Coordinate Models of the Point Cloud Data
		3.2 Orthographic Projection Grayscale Images Generation
		3.3 Distributed Parallel Data Processing
	4 Metro Shield Disease Detection Based on YOLOv5
		4.1 The Structure of YOLOv5
		4.2 Initial Anchor Box Optimization
		4.3 Loss Function Analysis and Optimization
	5 Experimental Results and Analysis
		5.1 Experimental Environment
		5.2 Performance Comparison Between the Single Sever and the Cluster
		5.3 Dataset of Metro Shield Diseases Construction
		5.4 Experiment and Analysis on Disease Detection Model
	6 Conclusions
	References
BlockFAD: A Federated Learning Based I/O Anomaly Detection Method for Blockchain Domain Name System
	1 Introduction
	2 Related Work
	3 Proposed Approach
		3.1 Overview
		3.2 Problem Formulation
		3.3 Data Preprocessing
		3.4 Two-channel Transformer Encoder
		3.5 Joint Optimization
	4 Experiment
		4.1 Dataset
		4.2 Evaluation Metrics
		4.3 Baseline
		4.4 Results and Analysis
		4.5 Ablation Experiments
	5 Conclusion
	References
Density and Viscosity Prediction of Mixtures Based on Multitasking and Attentional Mechanism
	1 Introduction
	2 Related Work
	3 Model Construction
		3.1 Model Architecture
		3.2 Parameter Sharing Layer
		3.3 Attention Mechanism Layer
		3.4 Multitask Layer
	4 Experiments and Analysis
		4.1 Experimental Data and Experimental Environment
		4.2 Evaluation Indicators
		4.3 Model Parameter Selection
		4.4 Comparative Experiment
	5 Conclusion
	References
Author Index