Neural Information Processing: 29th International Conference, ICONIP 2022, Virtual Event, November 22–26, 2022, Proceedings, Part VI (Communications in Computer and Information Science, 1793)

Preface
Organization
Contents – Part VI
Applications I
Transfer Learning Based Long Short-Term Memory Network for Financial Time Series Forecasting
	1 Introduction
	2 Basic Methodology
	3 Proposed Method
		3.1 Selection of Potential Source Dataset
		3.2 Explanation of ADA-LSTM
	4 Experiments
		4.1 Experimental Datasets
		4.2 Experimental Settings
		4.3 Empirical Results
		4.4 DM Test for Results
	5 Conclusion
	References
ScriptNet: A Two Stream CNN for Script Identification in Camera-Based Document Images*-12pt
	1 Introduction
	2 Proposed Methodology
		2.1 Spatial Stream
		2.2 Visual Stream
		2.3 Fusion of the Two Streams
	3 Experiments
		3.1 Dataset
		3.2 Implementation Details
		3.3 Ablation Study
		3.4 Comparison with Literature
	4 Conclusion
	References
Projected Entangled Pair State Tensor Network for Colour Image and Video Completion*-12pt
	1 Introduction
	2 Related Works
	3 PEPS Tensor Network
	4 Tensor Completion with PEPS Tensor Network
		4.1 Proposed Algorithm
		4.2 Complexity Analysis
	5 Numerical Experiments
		5.1 Synthetic Data
		5.2 Image Completion
		5.3 Video Completion
	6 Conclusion
	References
Artificial Neural Networks for Downbeat Estimation and Varying Tempo Induction in Music Signals
	1 Introduction
	2 Methods
		2.1 Recurrent Neural Networks
		2.2 Algorithm Description
		2.3 Evaluation Metrics and Procedure
		2.4 Data Sets
	3 Results and Discussion
		3.1 Ablation Study
		3.2 Comparative Analysis
	4 Conclusion and Future Work
	References
FedSpam: Privacy Preserving SMS Spam Prediction
	1 Introduction
	2 Literature Survey
	3 Methodology
		3.1 Dataset
		3.2 DistilBERT
		3.3 Federated Learning
		3.4 System Flow
	4 Experiment
		4.1 Experimental Setup
		4.2 Model Prediction Analysis
		4.3 On Device Resources Analysis
		4.4 Evaluation Metrics
	5 Results
	6 Conclusion and Future Scope
	References
Point Cloud Completion with Difference-Aware Point Voting
	1 Introduction
	2 Related Work
		2.1 Point Cloud Completion
		2.2 The Voting Mechanism
	3 Method
		3.1 Overview
		3.2 Seed Points Learning
		3.3 Hierarchical Point Decoder
	4 Experiments
		4.1 Datasets and Implementation Details
		4.2 Comparisons with State-of-the-Art Methods
		4.3 Ablation Studies
	5 Conclusion
	References
External Knowledge and Data Augmentation Enhanced Model for Chinese Short Text Matching
	1 Introduction
	2 Research Status and Background Knowledge
		2.1 Text Similarity Algorithms
		2.2 Data Augmentation
	3 Proposed Model
		3.1 Data Augmentation Model
		3.2 Input Model
		3.3 Semantic Information Transformer
		3.4 Sentence Matching Layer
		3.5 Relation Classifier Layer
	4 Experiments
		4.1 Datasets and Parameter Settings
		4.2 Compared Systems
		4.3 Main Results
		4.4 Analysis
	5 Conclusion
	References
Improving Oracle Bone Characters Recognition via A CycleGAN-Based Data Augmentation Method
	1 Introduction
	2 Related Works
		2.1 Oracle Bone Characters Recognition
		2.2 Generative Adversarial Networks (GANs)
	3 Method
		3.1 Statistics of OBC306
		3.2 Extraction of the Glyphs of Oracle Characters
		3.3 CycleGAN
		3.4 Samples Generation for Categories with Few Instances
	4 Experiments
		4.1 Experiments on Oracle Character Images Generation
		4.2 Experiments on Oracle Characters Recognition
	5 Conclusion
	References
Local-Global Interaction and Progressive Aggregation for Video Salient Object Detection
	1 Introduction
	2 Related Work
	3 The Proposed Method
		3.1 Overview of Network Architecture
		3.2 Symmetric RGB and Flow Streams
		3.3 Local-Global Interaction (LGI)
		3.4 Progressive Aggregation (PA)
		3.5 Loss Function
	4 Experiments
		4.1 Datasets and Metrics
		4.2 Implementation Details
		4.3 Comparisons to SOTAs
		4.4 Ablation Study
	5 Conclusion
	References
A Fast Stain Normalization Network for Cervical Papanicolaou Images
	1 Introduction
	2 Methodology
		2.1 Network Architecture
		2.2 Loss Function
	3 Experiments
		3.1 Normalization Evaluation
		3.2 Abnormal Cell Detection
		3.3 Ablation Studies
	4 Conclusion
	References
MEW: Evading Ownership Detection Against Deep Learning Models*-12pt
	1 Introduction
	2 Related Work
		2.1 Model Stealing
		2.2 Dataset Inference
	3 MEW
		3.1 Overview
		3.2 Inversion and Selection
		3.3 Fine-tune with EWC
	4 Experiments
		4.1 Dataset
		4.2 Experimental Setup
		4.3 Evaluation Results
	5 Discussion
	6 Conclusion
	References
Spatial-Temporal Graph Transformer for Skeleton-Based Sign Language Recognition*-12pt
	1 Introduction
	2 Proposed Method
		2.1 Preliminaries
		2.2 Model Overview
		2.3 Spatial Transformer
		2.4 Temporal Transformer
		2.5 Action and Continuous Sign Language Recognition
	3 Experiments
		3.1 Dataset
		3.2 Implementation Details
		3.3 Ablation Study
		3.4 Results on Action Recognition
		3.5 Results on Continuous Sign Language Recognition
	4 Conclusion
	References
Combining Traffic Assignment and Traffic Signal Control for Online Traffic Flow Optimization*-12pt
	1 Introduction
	2 Problem Definition
		2.1 Road Network
		2.2 Online Traffic Flow Optimization Problem
	3 Method
		3.1 Max-Throughput Control
		3.2 GEP-based Online Vehicle Navigation Algorithm
		3.3 Decision-Making Process
	4 Experiment
		4.1 Settings
		4.2 Compared Methods
		4.3 The Extracted Rule
		4.4 Experimental Results
	5 Conclusion
	References
Convolve with Wind: Parallelized Line Integral Convolutional Network for Ultra Short-term Wind Power Prediction of Multi-wind Turbines*-12pt
	1 Introduction
	2 Correlation Assumption and Spatio-temporal Analysis
		2.1 Time Delay
		2.2 Wake Effect
	3 Parallelized Line Integral Convolutional Network
		3.1 Line Feature Extraction
		3.2 Line Convolution
	4 Experiment Results and Discussions
		4.1 PLICN with Different Layers
		4.2 Contrast Experiment
	5 Conclusion and Prospects
	References
Bottom-Up Transformer Reasoning Network for Text-Image Retrieval
	1 Introduction
	2 Related Work
		2.1 Text-Image Retrieval
		2.2 Text-Image Retrieval Evaluation Metrics
	3 Bottom-Up Transformer Reasoning Network
		3.1 Image and Text Features Initialization
		3.2 Bottom-Up Transformer Reasoning Network (BTRN)
		3.3 Learning
		3.4 NDCG Metric for Text-Image Retrieval
	4 Efficiency
	5 Experiments
		5.1 Datasets and Protocols
		5.2 Implementation Details
		5.3 Experimental Result
		5.4 Qualitative Result and Analysis
	6 Ablation Study
		6.1 Averaging Versus Maxing
		6.2 Sharing Weights in Reasoning Phase
		6.3 Others
	7 Conclusion
	References
Graph Attention Mixup Transformer for Graph Classification
	1 Introduction
	2 Related Work
		2.1 Graph Transformers
		2.2 Mixup and Its Variants
	3 Methodology
		3.1 Graph Coarsening
		3.2 Graph Attention Mixup
		3.3 Graph Contrastive Learning with Mixup
	4 Experiments
		4.1 Experimental Setup
		4.2 Results and Analysis for Graph Classification
		4.3 Ablation Studies
		4.4 Sensitivity Analysis
	5 Conclusion
	References
Frequency Spectrum with Multi-head Attention for Face Forgery Detection*-12pt
	1 Introduction
	2 Related Work
		2.1 Datasets Generation Using GANs
		2.2 Spatial-Based Face Forgery Detection
		2.3 Frequency-Based Face Forgery Detection
		2.4 Vision Transformer(ViT)
	3 Proposed Methodology
		3.1 Architecture
		3.2 Two-Channel Extraction
		3.3 Multi-head Attention
	4 Experiment and Results
		4.1 Settings
		4.2 Results
		4.3 Ablation Study
	5 Conclusion and Future Work
	References
Autoencoder-Based Attribute Noise Handling Method for Medical Data
	1 Introduction
	2 Related Work
	3 A Method to Truly Handle Attribute Noise
	4 Experimental Results
		4.1 Used Datasets
		4.2 Used Metrics
		4.3 Experimental Protocol
		4.4 Results
	5 Conclusion
	References
A Machine-Reading-Comprehension Method for Named Entity Recognition in Legal Documents*-12pt
	1 Introduction
	2 NER as MRC
		2.1 Query Statement
		2.2 Dataset of Query, Context, and Answer
	3 BERT Fine-Tuning
	4 Extracting Answer Spans
		4.1 Biaffine Attention
		4.2 Add Rotary Position Embedding
		4.3 Train and Test
	5 Experiments
		5.1 Dataset
		5.2 Baselines
		5.3 Experimental Setting
		5.4 Results
		5.5 Case Study
	6 Related Work
		6.1 Nested Named Entity Recognition
		6.2 Machine Reading Comprehension
	7 Conclusion
	References
Cross-Modality Visible-Infrared Person Re-Identification with Multi-scale Attention and Part Aggregation*-12pt
	1 Introduction
	2 Related Work
	3 Method
		3.1 Overview
		3.2 Intra-modality Multi-scale Attention Module
		3.3 Fine-Grained Part Aggregation Learning
	4 Experiments
		4.1 Experimental Settings
		4.2 Comparison with State-of-Art Methods
		4.3 Ablation Study
	5 Conclusion
	References
Bearing Fault Diagnosis Based on Dynamic Convolution and Multi-scale Gradient Information Aggregation Under Variable Working Conditions
	1 Introduction
	2 Method
		2.1 Framework
		2.2 Dynamic Perception
		2.3 Gradient Aggregation
	3 Experiments and Analysis
		3.1 Introduction to the Dataset
		3.2 Result Analysis
		3.3 Comparison with Other Models
	4 Conclusions
	References
Automatic Language Identification for Celtic Texts
	1 Introduction
	2 Related Work
	3 Our Collected Dataset
	4 Our Approach
	5 Results
		5.1 Data Insights from the Unsupervised Models
		5.2 Supervised Experiments
		5.3 Semi-supervised Experiments
		5.4 Experiment with a Reduced Labelled Set
	6 Conclusions
	References
Span Detection for Kinematics Word Problems*-12pt
	1 Introduction
	2 Problem Definition
	3 Related Work
	4 JKS: Our Approach
		4.1 Span Detection Through Parameter-Enriched Representations
		4.2 Cross-Parameter Training of JKS
		4.3 Discussion
	5 Experiments and Results
		5.1 Datasets and Experimental Setup
		5.2 Evaluation Measure and Baselines
		5.3 Empirical Evaluation over hv and vp
		5.4 Empirical Evaluation over hv, vp and vv
		5.5 Discussion
	6 Conclusions and Future Work
	References
Emotion-Aided Multi-modal Personality Prediction System*-12pt
	1 Introduction
	2 Related Works
	3 Emotion-Enriched Personality Prediction Corpus
	4 The Proposed Approach
		4.1 Extraction of Personality Features
		4.2 Fusion of Personality and Emotion Features
		4.3 Personality Prediction
	5 Results and Discussion
		5.1 Implementation Details and Performance Evaluation
		5.2 Results
	6 Qualitative Analysis
	7 Conclusion and Future Works
	References
.25em plus .1em minus .1emKernel Inversed Pyramidal Resizing Network for Efficient Pavement Distress Recognition*-12pt
	1 Introduction
	2 Methodology
		2.1 Problem Formulation and Overview
		2.2 Image Resizing
		2.3 Pyramidal Convolution
		2.4 Kernel Inversed Convolution
		2.5 Multi-scale Pavement Distress Recognition
	3 Experiment
		3.1 Dataset and Setup
		3.2 Pavement Disease Recognition
		3.3 Ablation Study
	4 Conclusion
	References
Deep Global and Local Matching Network for Implicit Recommendation
	1 Introduction
	2 Related Work
	3 DeepGLM Model
		3.1 Global Matching Learning
		3.2 Local Matching Learning
		3.3 Gated Information Fusion
		3.4 Training Optimization
	4 Experiments
		4.1 Experimental Settings
		4.2 Overall Performance (RQ1)
		4.3 Effectiveness of Each Module (RQ2)
		4.4 Hyper-parameter Study (RQ3)
	5 Conclusion
	References
A Bi-hemisphere Capsule Network Model for Cross-Subject EEG Emotion Recognition*-12pt
	1 Introduction
	2 The Proposed Method
		2.1 The Bi-hemisphere Matrix
		2.2 Asymmetric Feature Learning
		2.3 EmotionCaps
		2.4 Discriminator
		2.5 The Entropy Loss
		2.6 The Optimization of Bi-CapsNet
	3 Experiments
		3.1 Datasets
		3.2 Implementation Details
		3.3 Results and Discussions
	4 Conclusion
	References
Attention 3D Fully Convolutional Neural Network for False Positive Reduction of Lung Nodule Detection*-12pt
	1 Introduction
	2 Methods
		2.1 U-Net with Squeeze and Excitation Blocks (U-SENet)
		2.2 Channel-Spatial Attention Based Fully CNN
		2.3 FC-C3D Network
	3 Experimental Results and Analysis
		3.1 Dataset
		3.2 Data Preprocessing
		3.3 Data Augmentation
		3.4 Candidate Nodules Extraction Based on U-SENet
		3.5 False Positive Reduction Based on FC-C3D Network
	4 Conclusions
	References
A Novel Optimized Context-Based Deep Architecture for Scene Parsing
	1 Introduction
	2 Related Works
		2.1 Optimization Algorithm
		2.2 Scene Parsing
	3 Proposed Method
		3.1 Visual Features
		3.2 Context Sensitive Features
		3.3 Optimization
	4 Results and Discussions
		4.1 Datasets
		4.2 Evaluation of Stanford Dataset
		4.3 Segmentation Results on CamVid Dataset
		4.4 Efficacy of the Proposed Optimization
	5 Conclusion
	References
.25em plus .1em minus .1emResnet-2D-ConvLSTM: A Means to Extract Features from Hyperspectral Image*-12pt
	1 Introduction
	2 Challenges and Objectives
	3 Model Description
	4 Data Set Description and Experimental Setup
	5 Result Analysis
	6 Conclusion
	References
An Application of MCDA Methods in Sustainable Information Systems
	1 Introduction
	2 Preliminaries
		2.1 The COMET Method
		2.2 The EDAS Method
		2.3 Rank Similarity Coefficients
	3 Practical Problem
	4 Results
	5 Conclusions
	References
Decision Support System for Sustainable Transport Development
	1 Introduction
	2 Interval TOPSIS Method
	3 Practical Problem
	4 Conclusions
	References
Image Anomaly Detection and Localization Using Masked Autoencoder*-12pt
	1 Introduction
	2 Related Work
	3 Approach
		3.1 Network Architecture
		3.2 Training and Inpainting Target
		3.3 Inference and Anomaly Detection
	4 Experiments
		4.1 Implementation Details
		4.2 Results and Discussion
	5 Ablation
	6 Conclusion
	References
Cross-domain Object Detection Model via Contrastive Learning with Style Transfer*-12pt
	1 Introduction
	2 Related Work
	3 Proposed Method
		3.1 Stylize Source Domain
		3.2 Contrastive Learning with InfoNCE
		3.3 Algorithm
	4 Experiments
		4.1 Datasets and Experiments Settings
		4.2 Use Cityscapes Datasets Adapt to Foggy Cityscapes Datasets
		4.3 Domain Transfer Using Other Datasets
		4.4 Ablation Experiments
		4.5 Qualitative Analysis
	5 Conclusion
	References
A Spatio-Temporal Event Data Augmentation Method for Dynamic Vision Sensor*-12pt
	1 Introduction
	2 Background
		2.1 DVS
		2.2 LSM
	3 Relatedwork
		3.1 Data Augmentation for Frame-based Data
		3.2 Data Augmentation for Event-based Data
	4 Motivation
	5 Method
		5.1 Random Translation
		5.2 Time Scaling
	6 Experiments and Results
		6.1 Datasets
		6.2 Experiment Setup
		6.3 Experiment Results
		6.4 Impact of Augmentation Level
	7 Conclusion
	References
FCFNet: A Network Fusing Color Features and Focal Loss for Diabetic Foot Ulcer Image Classification
	1 Introduction
	2 Related Work
	3 Proposed Method
		3.1 Preprocessing
		3.2 EfficientNet B3 and Focal Loss
		3.3 The Classification Network with the Segmentation Module
	4 Experiments
		4.1 Dataset
		4.2 Implementation Details
		4.3 Results and Analysis
		4.4 Ablation Study
		4.5 Some Deficiencies
	5 Conclusion and Future Work
	References
ClusterUDA: Latent Space Clustering in Unsupervised Domain Adaption for Pulmonary Nodule Detection
	1 Introduction
	2 Related Work
		2.1 PN Detection
		2.2 Domain Adaption
	3 Method
		3.1 Preliminary
		3.2 Motivation
		3.3 Cluster-based Domain Adaption
		3.4 PN Detection Network
		3.5 Network Overview
	4 Experiment
		4.1 Dataset
		4.2 Implementation Details
		4.3 Results
		4.4 Model Analysis
	5 Conclusion and Future Work
	References
Image Captioning with Local-Global Visual Interaction Network
	1 Introduction
	2 Related Work
	3 Proposed Approach
		3.1 Feature Extraction
		3.2 Visual Interaction Graph
		3.3 Relationship Attention Based LSTM
	4 Experiments
		4.1 Ablation Analysis
		4.2 Comparison with State-of-the-Arts
	5 Conclusion
	References
Rethinking Voxelization and Classification for 3D Object Detection
	1 Introduction
	2 Related Work
	3 Method
		3.1 Fast Dynamic Voxelizer
		3.2 RV Backbone Module
		3.3 Classification Sub-head Module
		3.4 Datasets
	4 Experimental Results
	5 Ablation Study
	6 Conclusions
	References
GhostVec: Directly Extracting Speaker Embedding from End-to-End Speech Recognition Model Using Adversarial Examples
	1 Introduction
	2 Background
		2.1 Transformer-Based E2E Speaker-Adapted ASR Systems
		2.2 Generating Adversarial Examples
	3 Proposed Methods
		3.1 GhostVec from Feature-Level Adversarial Examples
		3.2 GhostVec from Embedding-Level Adversarial Examples
	4 Experiment Setup
		4.1 Datasets
		4.2 ASR Models
		4.3 Evaluation Metrics
	5 Experimental Results
		5.1 Effectiveness of Adversarial Examples
		5.2 Effectiveness of GhostVec
		5.3 Further Discussions
	6 Conclusions
	References
An End-to-End Chinese and Japanese Bilingual Speech Recognition Systems with Shared Character Decomposition
	1 Introduction
	2 Related Work
		2.1 Multilingual ASR Systems
		2.2 The Decomposition Strategy for Chinese
	3 Proposed Method
		3.1 Decomposition Strategy
		3.2 Multilingual ASR Transformer
	4 Experiments
		4.1 Model Structure and Data Description
		4.2 Monolingual Baseline Systems
		4.3 Chinese-Japanese Bilingual Speech Recognition Task
	5 Conclusions and Future Work
	References
An Unsupervised Short- and Long-Term Mask Representation for Multivariate Time Series Anomaly Detection
	1 Introduction
	2 Related Work
	3 Methodology
		3.1 Mask and Sequence Split
		3.2 Multi-scale residual Convolution
		3.3 Joint Reconstruction and Forecasting Based on GRU
		3.4 Anomaly Detection
	4 Experiment and Analysis
		4.1 Datasets and Evaluation Metrics
		4.2 Performance and Analysis
		4.3 Ablation Study
		4.4 Localization Abnormal
	5 Conclusion
	References
Investigating Effective Domain Adaptation Method for Speaker Verification Task
	1 Introduction
	2 Proposed Method
	3 Experimental Settings
		3.1 Data
		3.2 Baseline System
		3.3 Domain Embedding
		3.4 Different Back-ends
		3.5 Data Augmentation
		3.6 Self-Supervised Learning
	4 Experimental Evaluations
		4.1 Experimental Results
		4.2 Further Discussions
	5 Conclusion
	References
Real-Time Inertial Foot-Ground Contact Detection Based on SVM
	1 Introduction
	2 Dataset
		2.1 Data Collection Equipment
		2.2 Data Collection
		2.3 Data Pre-processing
		2.4 Pressure Sensor Calibration
		2.5 Contact Label Generation
	3 Method
		3.1 Feature Engineering
		3.2 Machine Learning Models
		3.3 Experimental Setup
		3.4 Experimental Results
	4 Conclusion
	References
Channel Spatial Collaborative Attention Network for Fine-Grained Classification of Cervical Cells
	1 Introduction
	2 Materials and Methods
		2.1 Construction of LTCCD
		2.2 Channel Spatial Collaborative Attention Network
	3 Experiments
		3.1 Datasets
		3.2 Implementation Details and Evaluation Metrics
		3.3 Performance Comparison of Different CNN Architectures
		3.4 Performance Comparison Using Different Attention Modules
		3.5 Class Activation Visualization
	4 Conclusion and Future Works
	References
Multimodal Learning of Audio-Visual Speech Recognition with Liquid State Machine
	1 Introduction
	2 Background
		2.1 Liquid State Machine
		2.2 Dynamic Vision Sensor
		2.3 Related Works
	3 Methodology
		3.1 LSM Based Feature Extraction
		3.2 Fusion Function
		3.3 Recognition
	4 Experiments
		4.1 Experimental Setup
		4.2 Datasets and Preprocessing
		4.3 Training Strategy
		4.4 Experimental Results
	5 Conclusion
	References
Identification of Fake News: A Semantic Driven Technique for Transfer Domain
	1 Introduction
	2 Literature Review
	3 Theoretical Study
		3.1 Feature Selection Methods
		3.2 Machine Learning Techniques
	4 Proposed Method
		4.1 Fake News Datasets
		4.2 Evaluation Metrices
	5 Performance Evaluation and Analysis of Result
	6 Conclusion
	References
Portrait Matting Network with Essential Feature Mining and Fusion
	1 Introduction
	2 Related Work
		2.1 Image Matting
		2.2 End-to-End Portrait Matting
	3 Proposed Method
		3.1 Overview
		3.2 Critical Feature Extraction Block
		3.3 Context Fusion Block
		3.4 Detail Fusion Block
		3.5 Fusion Function
		3.6 Loss Function
	4 Experiments
		4.1 Implementation Details
		4.2 Dataset and Evaluation Metrics
		4.3 Ablation Study
		4.4 Experiment Results
	5 Conclusion
	References
Hybrid-Supervised Network for 3D Renal Tumor Segmentation in Abdominal CT
	1 Introduction
	2 Related Work
		2.1 Self-supervision
		2.2 Semi-supervision
	3 Approach
		3.1 Pre-training Stage
		3.2 Self-training Stage
	4 Experiments
	5 Conclusion
	References
Double Attention-Based Lightweight Network for Plant Pest Recognition
	1 Introduction
	2 Related Work
	3 Proposed Model
		3.1 The Lightweight Network
		3.2 Double Attention Layer
	4 Experiments
		4.1 Datasets
		4.2 Implementation and Training
		4.3 Results
		4.4 Ablative Study
	5 Conclusion
	References
A Deep Investigation of RNN and Self-attention for the Cyrillic-Traditional Mongolian Bidirectional Conversion
	1 Introduction
	2 Task Challenges
		2.1 Data Sparseness
		2.2 Agglutinative Characteristics
	3 RNN and Self-attention Based Encoder-Decoder Frameworks for C2T and T2C
		3.1 Overall Framework
		3.2 Workflow of C2T
		3.3 Workflow of T2C
		3.4 Backbone of Encoder-Decoder Framework
	4 Experiments and Results
		4.1 Datasets
		4.2 Evaluation Setup
		4.3 Performance Comparison for C2T and T2C
		4.4 Case Study
	5 Conclusion
	References
Sequential Recommendation Based on Multi-View Graph Neural Networks
	1 Introduction
	2 Related Work
	3 Method
		3.1 Interest Graph Construction
		3.2 Multi-view Attention Network
		3.3 Interest Extraction Layer
		3.4 Prediction Layer
	4 Experiment
		4.1 Datasets and Evaluation Metrics
		4.2 Parameter Setup
		4.3 Performance Comparison
		4.4 Ablation Experiments
	5 Conclusion
	References
Cross-Domain Reinforcement Learning for Sentiment Analysis
	1 Introduction
	2 Related Work
	3 Approach
		3.1 Overview of the Framework
		3.2 Cross-Domain Reinforcement Learning
		3.3 Feature Extraction
	4 Experiment
		4.1 Experimental Dataset and Settings
		4.2 Comparative Experiment
		4.3 Ablation Study
		4.4 Effect of Feature Selection Policy
	5 Conclusion and Future Work
	References
PPIR-Net: An Underwater Image Restoration Framework Using Physical Priors
	1 Introduction
	2 Related Work
	3 Proposed Method
		3.1 Structural Restoration Network
		3.2 Color Correction Network
		3.3 Loss Functions
	4 Experiments
		4.1 Datasets
		4.2 Evaluation Metrics
		4.3 Implementation Details
		4.4 Comparison with State-of-the-Art Methods
	5 Conclusion
	References
Denoising fMRI Message on Population Graph for Multi-site Disease Prediction
	1 Introduction
	2 Related Work
	3 Our Proposed Framework
		3.1 Problem Statement
		3.2 Proposed Framework
	4 Experiments
		4.1 Dataset and Preprocessing
		4.2 Baselines and Settings
		4.3 Main Results
		4.4 Ablation Study
		4.5 Visualization
	5 Conclusion
	References
CATM: Candidate-Aware Temporal Multi-head Self-attention News Recommendation Model
	1 Introduction
	2 Our Approach
		2.1 News Encoder
		2.2 Candidate-Aware User Encoder
		2.3 Click Prediction and Model Training
	3 Experiments
		3.1 Experiment Setup
		3.2 Performance Evaluation
		3.3 Ablation Studies
		3.4 Parameter Analysis
	4 Conclusion
	References
Variational Graph Embedding for Community Detection
	1 Introduction
	2 Related Work
		2.1 Community Detection and Node Representation Learning
		2.2 Joint Learning of Community Detection and Node Representation
	3 Proposed Method
		3.1 Problem Definition
		3.2 Graph Attention Networks
		3.3 Overall Framework
	4 Experiments
		4.1 Datasets
		4.2 Baselines
		4.3 Evaluation Metrics
		4.4 Parameter Settings
		4.5 Experimental Results
		4.6 Ablation Study
	5 Conclusion
	References
Counterfactual Causal Adversarial Networks for Domain Adaptation
	1 Introduction
	2 Related Works
		2.1 Unsupervised Domain Adaptation
		2.2 Causal Inference
	3 Methodology
		3.1 Counterfactual Causal Inference for Domain Adaptation
		3.2 Counterfactual Causal Adversarial Networks
		3.3 Objective Function
	4 Experiments
		4.1 Datasets and Setups
		4.2 Performance Analysis
	5 Conclusion
	References
Author Index