Artificial Neural Networks and Machine Learning - ICANN 2022

Name: Artificial Neural Networks and Machine Learning - ICANN 2022 | 31st International Conference on Artificial Neural Networks, Bristol, UK, September 6-9, 2022, Proceedings, Part II
Brand: Springer
Price: 53.49 EUR
Availability: OnlineOnly

31st International Conference on Artificial Neural Networks, Bristol, UK, September 6-9, 2022, Proceedings, Part II

Elias Pimenidis Plamen Angelov Chrisina Jayne Antonios Papaleonidas Mehmet Aydin(Editor)

Springer (Publisher)

Published on 6. September 2022

XXII, 813 pages

E-Book

PDF with digital watermarking

System requirements

978-3-031-15931-2 (ISBN)

€53.49incl. 7% vat

System requirements

for PDF with digital watermarking

E-Book Single Licence

Available for download

Description

More details

Other editions

Content

Intro
Preface
Organization
Contents - Part II
Alleviating Overconfident Failure Predictions via Masking Predictive Logits in Semantic Segmentation
1 Introduction
2 Related Work
3 Method
3.1 Preliminary
3.2 Masking Predictive Logits
3.3 Differences with Previous Methods
4 Experiments
4.1 Comparative Results
4.2 Qualitative Assessments
5 Conclusion
6 Appendix
6.1 Observing Overconfidence
6.2 Datasets and Metrics
6.3 Supplemental Experiments
References
Cooperative Multi-agent Reinforcement Learning with Hierachical Communication Architecture
1 Introduction
2 Related Work
3 Method
3.1 Communication Module
3.2 Sub-goal Generation Module
3.3 Action Generation Module
3.4 Learning
4 Experiments
4.1 Details
4.2 Results
4.3 Training Improvement
5 Conclusion and Futrue Work
References
Emotion Aware Reinforcement Network for Visual Storytelling
1 Introduction
2 Related Work
2.1 Visual Storytelling
3 Approach
3.1 Tracking Emotion in Text
3.2 Emotion Aware Storytelling Model
3.3 Training
4 Experiments
4.1 Dataset
4.2 Quantitative Evaluation
4.3 Ablation Study
4.4 Effect of the Hyper-parameter , and
4.5 Human Evaluation
4.6 Implementation Details
5 Conclusion
References
Long-Horizon Route-Constrained Policy for Learning Continuous Control Without Exploration
1 Introduction
2 Related Work
3 Problem Statement
4 Long-horizon Route-constrained Policy
5 Experiments
5.1 Dataset
5.2 Implementation Details
5.3 Results
6 Conclusion
References
Model-Based Offline Adaptive Policy Optimization with Episodic Memory
1 Introduction
2 Related Work
2.1 Model-Based Offline RL
2.2 Episodic Memory-Based Methods
3 Preliminaries
4 Approach
4.1 Adaptive Constraint Strength Optimization
4.2 Episodic Memory
4.3 Practical Implementation
5 Experiment
5.1 Experimental Dataset and Settings
5.2 Comparative Methods
5.3 Comparative Experiment
5.4 Evaluation on Tasks Requiring OOD Generalization
5.5 Ablation Study
6 Conclusion and Future Work
References
Multi-mode Light: Learning Special Collaboration Patterns for Traffic Signal Control
1 Introduction
2 Related Works
2.1 Conventional Traffic Control
2.2 Reinforcement Learning
3 Problem Definition
4 Method
4.1 Observation Embedding Layer
4.2 Adaptive Neighborhood Cooperation Layer
4.3 Q-value Prediction
5 Experiment
5.1 Settings
5.2 Datasets
5.3 Compared Method
5.4 Evaluation Metric
5.5 Performance Comparison
5.6 Ablation Experiments
6 Conclusion
References
Pheromone-inspired Communication Framework for Large-scale Multi-agent Reinforcement Learning
1 Introduction
2 Related Work
2.1 Multi-agent Reinforcement Learning
2.2 Control Algorithms Based on Swarm Intelligence
3 Background
3.1 Problem Formulation
3.2 Tabular Q-Learning and Deep Q-Learning
4 Methodology
4.1 Communication Framework Overview
4.2 Virtual Medium Settings
4.3 Pheromone Update Rule
4.4 Pheromone Inspired Q-Learning
5 Experiment
5.1 Main Experiments
5.2 Detailed Experiments in Battle
6 Conclusion
References
Reinforcement Learning for the Pickup and Delivery Problem
1 Introduction
2 Related Work
3 Problem Definition
4 Reinforcement Learning Model
4.1 Actor Network
4.2 Critic Network
4.3 Update and Mask
5 Experiments
5.1 Experiment Setting
5.2 Baselines
5.3 Results
5.4 Training Pace
6 Conclusions
References
Towards Relational Multi-Agent Reinforcement Learning via Inductive Logic Programming
1 Introduction
2 Related Work
3 Preliminary
3.1 Inductive Logic Programming
3.2 Differentiable Inductive Logic Programming
4 Methodology
4.1 Problem Setting
4.2 Partner-Aware Actor-Critic Gradient Policy
4.3 Multi-agent Reinforcement Learning with DILP
5 Experimental Results
5.1 Experiment Setup
5.2 Results and Analysis
6 Conclusion
References
Understanding Reinforcement Learning Based Localisation as a Probabilistic Inference Algorithm
1 Introduction
2 Related Work
3 Method
3.1 Problem Setting
3.2 MDP Setting
3.3 Reward Function
4 Connection to Inference Algorithm on Graphical Model
4.1 Connection to CRF
4.2 Connection to HMM and MEMM
5 Evaluation
5.1 SPHERE Challenge Dataset
5.2 Evaluation Procedure
5.3 Baseline Algorithms
5.4 Results
6 Conclusions
References
Word-by-Word Generation of Visual Dialog Using Reinforcement Learning
1 Introduction
1.1 Task Definition
2 Related Work
2.1 Visual Dialog Generation
2.2 Compositionality in VQA
3 Methodology
3.1 Data Pre-processing
3.2 Architecture
3.3 Loss Function
4 Experiments
4.1 Pre-training
4.2 Experiments with Different Time Efficiency Losses
5 Results
5.1 Accuracy
5.2 Question Evaluation
6 Conclusion
References
A Novel Approach to Train Diverse Types of Language Models for Health Mention Classification of Tweets
1 Introduction
2 Related Work
2.1 Adversarial Training
2.2 Self-supervised Representation Learning
2.3 Health Mention Classification of Tweets
3 Methodology
3.1 Adversarial Training
3.2 Barlow Twins (BT) Loss
3.3 Adversarial Training with BT Loss for HMC
4 Experiments
4.1 Dataset
4.2 Experimental Setup
5 Results and Analysis
5.1 Effect of Noise on Layers Level of Models
5.2 Effect of Noise Amount on Model's Performance
5.3 Comparison with SOTA
5.4 Explaining the Model Decision
6 Conclusion
References
Adaptive Knowledge Distillation for Efficient Relation Classification
1 Introduction
2 Related Work
2.1 Relation Classification
2.2 Knowledge Distillation
3 Methods
3.1 Confusion-Based Dynamic Temperature Distillation
3.2 Logit Replacement for Knowledge Adjustment
3.3 Combination
4 Experiment Results and Discussions
4.1 Dataset and Evaluation Metrics
4.2 Experimental Setting
4.3 Model Performance Comparison
4.4 Model Efficiency Comparison
5 Conclusions
References
An Adversarial Multi-task Learning Method for Chinese Text Correction with Semantic Detection
1 Introduction
1.1 Bottlenecks and Defects
1.2 Motivation and Contributions
1.3 Achievements
2 Methodology
2.1 Task Formulation
2.2 Adversarial Multi-task Learning
2.3 Chinese Text Correction with Semantic Error Detection
3 Experimental Result
3.1 Dataset
3.2 Training Settings
3.3 Ablation Results
3.4 Main Results
4 Conclusion
References
An Unsupervised Sentence Embedding Method by Maximizing the Mutual Information of Augmented Text Representations
1 Introduction
2 Related Work
2.1 Unsupervised Sentence Embedding
2.2 Mutual Information and Representation Learning
3 Model
3.1 Description
3.2 Analysis
3.3 Model Architecture
4 Experiment
4.1 Dataset and Setups
4.2 Results
5 Qualitative Analysis
5.1 Analysis of Multi-head Attention
5.2 Analysis of Dropout
6 Ablation Study
6.1 Dropout for Global Embeddings
6.2 Combination Experiments
7 Conclusion
References
Analysis of COVID-19 5G Conspiracy Theory Tweets Using SentenceBERT Embedding
1 Introduction
2 Related Work
3 Problem Definition and Challenges
4 Dataset
5 Features
5.1 Embedding
5.2 External Features
6 Classification
7 Results
8 Conclusion
References
Chinese Named Entity Recognition Using the Improved Transformer Encoder and the Lexicon Adapter
1 Introduction
2 Background
2.1 Self-attention
2.2 Absolute Position Embedding
3 Related Work
4 Our Model
4.1 Attention Adapter for Character-word Fusion
4.2 Our Relative Position Embedding
4.3 Our Attention Algorithm
5 Experiments
5.1 Datasets
5.2 Experimental Settings
5.3 Comparison with Previous Models
6 Conclusion and Future Work
References
Concatenating BioMed-Transformers to Tackle Long Medical Documents and to Improve the Prediction of Tail-End Labels
1 Introduction
2 Related Work
3 Data and Labels
4 Language Models
5 Concatenated Language Models
6 Experiments
7 Results
7.1 Overall Performance
7.2 SOTA Results
7.3 Tail-End Labels
8 Discussion
References
Eliciting Knowledge from Pretrained Language Models for Prototypical Prompt Verbalizer
1 Introduction
2 Related Work
3 Method
3.1 Overview
3.2 Few-Shot Settings
3.3 Zero-Shot Settings
4 Experiments
4.1 Datasets and Templates
4.2 Baselines
4.3 Implementation Details
4.4 Results and Analysis
5 Conclusion
References
Integrating Label Semantic Similarity Scores into Multi-label Text Classification
1 Introduction
2 Related Work
3 Methodology
3.1 Problem Definition
3.2 One-Hot Vector Scoring Network
3.3 Label Similarity Scoring Network
3.4 Adaptive Weighted Loss
4 Experiment Setup
4.1 Datasets
4.2 Baseline Models
4.3 Parameter Setting
4.4 Evaluation Metrics
5 Experimental Results
5.1 Comparison Results and Discussion
5.2 Ablation Test
5.3 Comparison on Tail Labels
5.4 Different Hyperparameters of AWLoss
5.5 Applications in COVID-19 Literature Multi-label Classification
6 Conclusion
References
Learning Flexible Translation Between Robot Actions and Language Descriptions
1 Introduction
2 Related Work
3 Proposed Method: PGAE
4 Experiment Results
5 Conclusion
References
Learning Visually Grounded Human-Robot Dialog in a Hybrid Neural Architecture
1 Introduction
2 Background and Related Work
3 Multimodal Dataset for Human-Robot Interaction
3.1 Human-Robot Conversation Task Definition
3.2 Visual Scenes
3.3 Conversations
4 Approach
4.1 Object Detection (OD)
4.2 Dialog State Tracker (DST)
4.3 Human-Robot Interaction Policy (HRIP)
5 Experiments
5.1 Results
5.2 Discussion
6 Conclusion and Future Work
References
MTHGAT: A Neural Multi-task Model for Aspect Category Detection and Aspect Term Sentiment Analysis on Restaurant Reviews
1 Introduction
2 Related Work
3 Multi-task Hierarchical Graph Attention Network
3.1 Graph Attention Network
3.2 Input-Embedding Module
3.3 Sentence-level GAT Module
3.4 Convolution and Pooling Module
3.5 Document-level GAT Module
3.6 Softmax Module for ACD
3.7 Softmax Module for ATSA
3.8 Model Training
4 Experimentation
4.1 Datasets and Experimental Settings
4.2 Baseline Methods
4.3 Experimental Results
4.4 Ablation Study
4.5 Error Analysis
5 Conclusion
References
Multi-task Alignment Scheme for Span-level Aspect Sentiment Triplet Extraction
1 Introduction
2 Related Work
3 Proposed Model
3.1 Initial Encoding Module
3.2 Aspect/Opinion Semantic Composition Module
3.3 Pointer-specific Tagging Module
3.4 Triplet Alignment Module
3.5 Training Loss
4 Experiments
4.1 Datasets and Experimental Setting
4.2 Baselines
4.3 Experimental Results
4.4 Ablation Study
4.5 Case Study and Error Analysis
5 Conclusion
References
SSMFRP: Semantic Similarity Model for Relation Prediction in KBQA Based on Pre-trained Models
1 Introduction
2 Related Work
3 Approach
3.1 Overview
3.2 Sentence Pair Generation
3.3 Score Calculation
4 Experiments
4.1 Dataset
4.2 Experiment Setting
4.3 Experimental Results
4.4 Ablation Experiments
5 Conclusion
References
SubCrime: Counterfactual Data Augmentation for Target Sentiment Analysis
1 Introduction
2 Related Work
3 SubCrime Model
3.1 Step 1: Substitute Target Words
3.2 Step 2: Discriminate Synthesized Sentence
4 Experiments
4.1 Datasets
4.2 Evaluation Metrics and Settings
4.3 Experimental Results
4.4 Case Study
5 Conclusion
References
Word Embeddings with Fuzzy Ontology Reasoning for Feature Learning in Aspect Sentiment Analysis
1 Introduction
2 The Proposed Aspect-Based Sentiment Analysis Approach
2.1 Pre-processing
2.2 Feature Polarity Identification
2.3 Feature Extraction
2.4 Sentiment Classification
3 Experimental Results and Discussion
3.1 Experimental Results
3.2 Comparative Analysis
4 Conclusions and Future Work
References
3D Face Reconstruction with Geometry Details from a Single Color Image Under Occluded Scenes
1 Introduction
2 Related Work
2.1 Single Image 3D Face Reconstruction
2.2 Generative Adversarial Networks
2.3 Face Image Synthesis
3 Proposed Approach
3.1 Face Parsing Map Generation
3.2 Face Edge Lines Map Generation
3.3 Recovering 3D Face Geometric Details
4 Implementation Details
5 Experimental Results
5.1 Qualitative Comparisons with Recent Art
5.2 Quantitative Comparison with Recent Art
6 Conclusions
References
A Transformer-Based GAN for Anomaly Detection
1 Introduction
2 Proposed Method
2.1 Model Overview
2.2 U-Shape Generator
2.3 Swin Transformer Block
2.4 Skip Attention Connection
2.5 Loss Function
2.6 Inference
3 Experiment
3.1 Dataset
3.2 Training Details
3.3 Encoder of Our Transformer-Based Structure
3.4 Experimental Analysis
4 Conclusion
References
AMMUNIT: An Attention-Based Multimodal Multi-domain UNsupervised Image-to-Image Translation Framework
1 Related Work
2 Methods
2.1 Preliminaries
2.2 Framework Architecture
2.3 Training Objectives
3 Experiments
4 Results
5 Conclusions and Discussion
References
Continual Learning by Task-Wise Shared Hidden Representation Alignment
1 Introduction
2 Related Works
3 Task-Wise Shared Hidden Representation Alignment (TSHRA)
3.1 Problem Setting
3.2 Our Proposed Model
4 Experiments
4.1 Datasets
4.2 The CL Baseline
4.3 Experimental Results and Comparison
5 Conclusions and Future Work
References
Contrast and Aggregation Network for Generalized Zero-shot Learning
1 Introduction
2 Related Work
2.1 Embedding Methods
2.2 Generative Methods
3 Methodology
3.1 Problem Setting
3.2 Visual Feature Generation
3.3 Feature Contrast and Aggregation
3.4 Semantic Feature Alignment
4 Experiments
4.1 Datasets
4.2 Evaluation Protocols
4.3 Implementation Details
4.4 Performance on Generalize Zero-shot Learning
5 Conclusion
References
DT2I: Dense Text-to-Image Generation from Region Descriptions*6pt
1 Introduction
2 Related Work
3 Method
3.1 Dense-Text-Conditional GAN
3.2 Regional Semantic Image-Text Matching
3.3 Training Objectives and Implementation Details
4 Experiments
4.1 Synthetic Images
4.2 Real Images
5 Conclusion
References
Image Inpainting Based Multi-scale Gated Convolution and Attention
1 Introduction
2 Related Work
2.1 Image Inpainting
2.2 Gated Convolution
3 Method
3.1 Inpainting Network
3.2 Multi-scale Gated Convolution
3.3 Scale Attention Block
3.4 Loss Functions
4 Experiments
4.1 Experiments Settings
4.2 Qualitative Comparisons
4.3 Quantitative Comparisons
4.4 Ablation Stud
4.5 Additional Results
5 Conclusion
References
Pancreatic Image Augmentation Based on Local Region Texture Synthesis for Tumor Segmentation
1 Introduction
2 Proposed Method
2.1 The Overall Framework of TSTG-GAN
2.2 The Process of Fusing the New Lesion
2.3 Optimization Objectives of TSTG-GAN
2.4 TSTG-GAN Network and Application of Synthetic Samples
3 Experiments and Results
3.1 Datasets and Pre-processing
3.2 Comparison Methods and Evaluation Indicators
3.3 Experimental Results and Analysis
4 Discussions and Conclusions
References
Phenotype Anomaly Detection for Biological Dynamics Data Using a Deep Generative Model
1 Introduction
2 Biological Time Series Data of Cell Division
2.1 Gene Function Identification Through Knocking-Down a Gene
2.2 Developmental Dynamics Data of C. elegans as a model animal
2.3 Representation of Three-Dimensional Nucleus Data
3 Anomaly Detection by Variational Auto-Encoder
3.1 Auto-encoder and Variational Auto-Encoder
3.2 Morphological Anomaly Detection by a Reconstruction Error
3.3 Temporal Anomaly Detection by a Latent Space Position
4 Computational Experiments of Morphological and Temporal Anomaly Detection on RNAi Embryos
4.1 Anomaly Detection by VAE Using a Set of Cross Section Data
4.2 Anomaly Detection by AE Using Voxel Data
5 Summary
References
Progressive Image Restoration with Multi-stage Optimization
1 Introduction
2 Method
2.1 Generator
2.2 Discriminator
2.3 GAA
2.4 Loss Function
3 Experiments
3.1 Training Setting and Strategy
3.2 Datasets
3.3 Comparison Models
4 Results
4.1 Qualitative Comparison
4.2 Quantitative Comparisons
4.3 Ablation Studies
5 Conclusion
References
A Unified View on Self-Organizing Maps (SOMs) and Stochastic Neighbor Embedding (SNE)
1 Introduction
1.1 Notations and Definitions
1.2 SOMs in This Framework
1.3 SNE in This Framework
2 Quantitative Comparisons
2.1 Comparison Approach
2.2 Toy 2D Datasets
2.3 Fashion-MNIST Dataset
3 Conclusion
References
Decoupled Representation Network for Skeleton-Based Hand Gesture Recognition
1 Introduction
2 Related Works
2.1 Skeleton Representations
2.2 Deep Neural Networks
3 Our Approach
3.1 Overview
3.2 Temporal Perception Branch
3.3 Spatial Perception Branch
4 Experiments
4.1 Datasets
4.2 Implementation Details
4.3 Ablation Study
4.4 Comparison with the State-of-the-Art
5 Conclusion
References
Dual Perspective Contrastive Learning Based Subgraph Anomaly Detection on Attributed Networks
1 Introduction
2 Related Work
2.1 Shallow Mechanism Method
2.2 Deep Learning Method
2.3 Contrastive Learning Method
3 Methods
3.1 Problem Formulation
3.2 Probability Sampling Module
3.3 Contrastive Learning Module
3.4 Dual Perspective Inference Module
4 Experiments
4.1 Datasets
4.2 Experimental Setup
4.3 Anomaly Detection Results and Analysis
4.4 Hyperparameter Study
4.5 Ablation Study
5 Conclusion
References
GCMK: Detecting Spam Movie Review Based on Graph Convolutional Network Embedding Movie Background Knowledge
1 Introduction
1.1 Background
1.2 Challenges
1.3 Contributions
2 Related Work
3 Methodology
3.1 Dataset Construction
3.2 Detection Model
3.3 Feature Concatenation
4 Experiments
4.1 Experiment Settings
4.2 Baseline Model Comparison Experiment
4.3 Feature Ablation Experiment
4.4 Robustness Experiment
5 Conclusion
References
Heterogeneous Graph Attention Network for Malicious Domain Detection
1 Introduction
2 Related Work
3 Preliminary
4 The System Description of HANDom
4.1 Data Preprocessing
4.2 HIN Construction
4.3 Graph Pruning
4.4 Meta-path Based Neighbors Extraction
4.5 HAN Classification
5 Experiments
5.1 Dataset
5.2 Comparative Methods
5.3 Experimental Setting
5.4 Performance Comparisons
5.5 Ablation Study
5.6 Parameter Analysis
6 Conclusion
References
IA-ICGCN: Integrating Prior Knowledge via Intra-event Association and Inter-event Causality for Chinese Causal Event Extraction
1 Introduction
2 Related Work
2.1 Pattern Matching for CEE
2.2 Statistical Learning for CEE
2.3 Neural Networks for CEE
3 Our Model
3.1 Feature Encoder
3.2 Intra-event Association and Inter-event Causality Networks Construction
3.3 Prior Knowledge Encodeing with GCN
3.4 Causality Extracting with BiLSTM+CRF
4 Experiment
4.1 Datasets
4.2 Experimental Setting
4.3 Baseline Methods
4.4 Comparison with Baseline Methods
4.5 Comparison w.r.t. Epoch
4.6 Ablation Experiments
5 Conclusion
References
Knowledge Graph Bidirectional Interaction Graph Convolutional Network for Recommendation
1 Introduction
2 Related Work
3 Methodology
3.1 Problem Formulation
3.2 Embedded Layer
3.3 Knowledge Graph Awareness Propagation Layer
3.4 BI-Interaction Aggregation Layer
3.5 Predict Layer
3.6 Optimization
4 Experiments
4.1 Datasets
4.2 Baselines
4.3 Experiments Settings
4.4 Results
5 Conclusions
References
Learning Hierarchical Graph Convolutional Neural Network for Object Navigation
1 Introduction
2 Related Work
3 Model
3.1 Task Definition
3.2 Model Overview
3.3 HGCNN
4 Experiments
4.1 Environment Settings
4.2 Evaluation
4.3 Implementation Details
4.4 Ablation Study
4.5 Quantitative Results
5 Conclusion
References
Low-Level Graph Convolution Network for Point Cloud Processing
1 Introduction
2 Related Work
3 Graph on Point Clouds
3.1 Background
3.2 Our Low-Level Graph Convolution
3.3 Network Architecture
4 Experiments
4.1 Shape Classification
4.2 Shape Part Segmentation
5 Ablation Study
6 Complexity Analysis
7 Conclusion
References
Low-Resource Similar Case Matching in Legal Domain
1 Introduction
2 Related Work
3 Methodology
3.1 Data Augmentation
3.2 Model
3.3 Training and Inference
4 Experiments
4.1 Setup
4.2 Performance Evaluation
4.3 Effect of Data Augmentation
4.4 Effect of Global Information
5 Conclusion and Future Work
References
Message Passing Neural Networks for Hypergraphs
1 Introduction
2 Related Work
2.1 Graph Design Space
2.2 Hypergraph Neural Networks and Hypergraph Representation
3 Graph Expansions and Loss of Structural Information
4 The Proposed Hypergraph Message Passing Neural Networks
4.1 Batch Normalization and Dropout
4.2 Equivalency Between Hypergraph Convolutions and HMPNN
5 Experiments
5.1 Experiment Setup
5.2 Results and Analysis
6 Conclusion and Future Work
References
Multiscale Spatial and Temporal Learning for Human Motion Prediction
1 Introduction
2 Related Work
2.1 Human Motion Prediction
2.2 GCN Based Spatial Relationship Modeling
3 Our Approach
3.1 Human Motion Encoding Module
3.2 Spatio-Temporal Graph Convolution Module
3.3 Multiscale Temporal Encoding Module
3.4 Loss Function
4 Experiments
4.1 Dataset and Preprocessing
4.2 Implementation Details
4.3 Results
4.4 Ablation Study
5 Conclusion
References
Multi-view Cascading Spatial-Temporal Graph Neural Network for Traffic Flow Forecasting
1 Introduction
2 Related Work
2.1 Machine Learning and Convolutional Method
2.2 Graph Method
2.3 Traffic Flow Prediction
3 Problem Formulation
4 Model
4.1 Differencing Process
4.2 Graph
4.3 Graph Layer
4.4 Graph Module and Graph Block
4.5 Attention Module
5 Experiments
5.1 Data Preparation
5.2 Experiment Settings
5.3 Experiment Result
6 Ablation Study
7 Conclusion
References
Parallel Message Passing in Dual-space on Graphs
1 Introduction
2 Related Work
3 Problem
4 The Framework
4.1 Overview
4.2 Parallel Message Passing Module
4.3 Properties of PSGC
4.4 Discussion
5 Experiments
5.1 Datasets
5.2 Experimental Settings
5.3 Overall Results
5.4 Over-Smoothing Analysis
6 Conclusion
References
Parallel Relationship Graph to Improve Multi-Document Summarization
1 Introduction
2 Related Work
2.1 Abstractive MDS
2.2 Graph-based Summarization
3 Model
3.1 Document Encoder
3.2 Graph Encoder
3.3 Graph Decoder
4 Experiments
4.1 Datasets and Experimental Settings
4.2 Baseline Models
4.3 Results
5 Conclusion
References
SNNet: Specific Node Network of Human Parsing
1 Introduction
2 Related Works
2.1 Human Parsing
2.2 Graph Reasoning
3 Specific Node Network
3.1 Overall Framework
3.2 Node Extract Module
3.3 Low-High Module
3.4 Multi-set Union Module
3.5 Loss Function
4 Experiments
4.1 Datasets
4.2 Implementation Details
4.3 Ablation Study
4.4 Comparison of State of Art
5 Conclusion
References
Spatial-Temporal Attention Network for Crime Prediction with Adaptive Graph Learning
1 Introduction
2 Related Work
3 Problem Formulation
4 Methodology
4.1 Architecture of AGL-STAN
4.2 Adaptive Graph Learning Module
4.3 Time-Aware Self-attention Module
4.4 Optimization
5 Experiments
5.1 Methodology
5.2 Prediction Accuracy
5.3 Computational Cost
5.4 Model Ablation Study
5.5 Model Effectiveness
5.6 Model Hyper-Parameters
6 Conclusion
References
ST2PE: Spatial and Temporal Transformer for Pose Estimation
1 Introduction
2 Related Work
2.1 End-to-End 3D Pose Estimation with Image
2.2 Two-Stage 3D Pose Estimation with Image
2.3 3D Pose Estimation with Video
2.4 Multi-view 3D Pose Estimation
3 Method
3.1 Single-View Spatial and Temporal Transformer
3.2 Multi-view Spatial and Temporal Transformer for Pose Estimation
4 Experiment
4.1 Datasets and Evaluation Metrics
4.2 Implementation Details
4.3 Comparison with State-of-the-Art
4.4 Ablation Study
4.5 Generalization Performance
5 Conclusion
References
Taking Care of Our Drinking Water: Dealing with Sensor Faults in Water Distribution Networks
1 Introduction
2 Related Work
3 Water Networks
4 Method
4.1 Fault Detection
4.2 Fault Isolation
4.3 Fault Accommodation
5 Experiments
5.1 Data
5.2 Fault Detection
5.3 Fault Isolation
5.4 Fault Accommodation
6 Conclusion
References
Temporal Graph Transformer for Dynamic Network
1 Introduction
2 Related Work
3 Methodology
3.1 Preliminary
3.2 Overall Structure
3.3 Update Module
3.4 Aggregation Module
3.5 Propagation Module
4 Experiments
4.1 Dataset
4.2 Experimental Setting
4.3 Link Prediction
4.4 Edge Classification
4.5 Ablation Study
4.6 Efficiency and Robustness Analysis
5 Conclusion
References
Towards Understanding the Effect of Node Features on the Predictions of Graph Neural Networks
1 Introduction
2 Related Works
3 Methods
3.1 Problem Formulation
3.2 GNN Prediction Interpreter
4 Experiment
4.1 Experiment Settings
4.2 Explanation Results
5 Conclusion
References
Weight-Aware Graph Contrastive Learning
1 Introduction
2 Related Works
3 Methods
3.1 Data Augmentation
3.2 Edge-Weight Sensitive Graph Attentional Layers
3.3 Weight-Aware Loss
4 Experiments
4.1 Comparison with State-of-the-Art Methods
4.2 Comparison Between Different Encoders
4.3 Visual Analysis
5 Conclusions
References
Why Deeper Graph Neural Network Performs Worse? Discussion and Improvement About Deep GNNs
1 Introduction
2 Preliminaries
3 Related Work and Drawback of GNNs
4 Proposed Method
4.1 Derivation of the Proposed Method Using SiameseNet
4.2 Pseudocode of the Proposed Method
4.3 Theoretical Discussion of the Proposed Method
5 Experimental Evaluation
5.1 Experimental Setting
5.2 Experimental Results
6 Conclusion
References
A Long and Short Term Preference Model for Next Point of Interest Recommendation
1 Introduction
2 Related Work
3 Method
3.1 Long-Term Preference Model
3.2 Short-Term Preference Model
3.3 Recommendation Combined Long-Term and Short-Term Preference Model
3.4 Model Training
4 Experiments
4.1 Datasets
4.2 Parameter Settings
4.3 Results and Analysis
5 Conclusion
References
A Multi-stack Denoising Autoencoder for QoS Prediction
1 Introduction
2 QoS Prediction with MSDAE
2.1 Pre-populated
2.2 Get Users' Preferences
2.3 QoS Prediction
3 Experiment Analysis
3.1 Dataset
3.2 Evaluation Indicators
3.3 Parameter Settings
3.4 Baseline Comparison
4 Summary and Future Work
References
CKEN: Collaborative Knowledge-Aware Enhanced Network for Recommender Systems
1 Introduction
2 Problem Formulation
3 Method
3.1 Collaborative Propagation Layer
3.2 Knowledge Graph Propagation Layer
3.3 Feature Interaction Layer
3.4 Prediction Layer
3.5 Model Learning
4 Experiments
4.1 Dataset
4.2 Baselines
4.3 Experiments Setup
4.4 Performance Comparison
4.5 Study of CKEN
5 Conclusion and Future Work
References
Conditioned Variational Autoencoder for Top-N Item Recommendation
1 Introduction
2 Background
2.1 Notation
2.2 Variational Autoencoder
3 Related Works
4 Conditioned Variational Autoencoder
4.1 Architecture
4.2 Conditioned Loss Function
5 Experiments
5.1 Datasets
5.2 Conditions Computation
5.3 Experimental Setup
5.4 Experimental Results and Discussion
5.5 Analysis of the C-VAE Produced Rankings
6 Conclusions and Future Work
References
Personalized Headline Generation with Enhanced User Interest Perception
1 Introduction
2 Related Work
2.1 News Recommendation Module
2.2 News Headline Generation Module
2.3 User Interest Embedding Module
3 Methods
3.1 User Interest Representation Using Entity Word Enhancement
3.2 User Interest Dynamic Filtering Module to Enhance User Interest Perception
3.3 User Interest Injection Module for Personalization
4 Experiments
4.1 Datasets
4.2 Implementation Details
4.3 Evaluation Metrics
4.4 Compared Methods
4.5 Experimental Results
4.6 Ablation Study
5 Conclusions
References
Correction to: A Multi-stack Denoising Autoencoder for QoS Prediction
Correction to: Chapter "A Multi-stack Denoising Autoencoder for QoS Prediction" in: E. Pimenidis et al. (Eds.): Artificial Neural Networks and Machine Learning - ICANN 2022, LNCS 13530, https://doi.org/10.1007/978-3-031-15931-2_62
Correction to: Artificial Neural Networks and Machine Learning - ICANN 2022
Correction to: E. Pimenidis et al. (Eds.): Artificial Neural Networks and Machine Learning - ICANN 2022, LNCS 13530, https://doi.org/10.1007/978-3-031-15931-2
Author Index

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Artificial Neural Networks and Machine Learning - ICANN 2022

Description

More details

Other editions

Additional editions

Content

System requirements