Advances in Knowledge Discovery and Data Mining

Name: Advances in Knowledge Discovery and Data Mining | 28th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2024, Taipei, Taiwan, May 7-10, 2024, Proceedings, Part IV
Brand: Springer
Price: 139.09 EUR
Availability: OnlineOnly

28th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2024, Taipei, Taiwan, May 7-10, 2024, Proceedings, Part IV

De-Nian Yang Xing Xie Vincent S. Tseng Jian Pei Jen-Wei Huang Jerry Chun-Wei Lin(Herausgeber*in)

Springer (Verlag)

Erschienen am 30. April 2024

XXXIII, 352 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-981-97-2238-9 (ISBN)

139,09 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Intro
General Chairs' Preface
PC Chairs' Preface
Organization
Contents - Part IV
Financial Data
Look Around! A Neighbor Relation Graph Learning Framework for Real Estate Appraisal
1 Introduction
2 Preliminaries
3 Approach
3.1 Graph Construction
3.2 Transaction Encoding
3.3 Neighbor Aggregator
3.4 Community Aggregator
3.5 Dynamic Adaptor
4 Experiments
4.1 Experimental Setup
4.2 Overall Performance
4.3 Ablation Study
5 Conclusion
References
Multi-time Window Ensemble and Maximization of Expected Return for Stock Movement Prediction
1 Introduction
2 Related Works
3 Proposed Model
3.1 Problem Formulation
3.2 Multi-time Window Ensemble Classifier
3.3 Base Learner
3.4 Proposed Loss Function for Base Learner
4 Experiments
4.1 Experimental Setup
4.2 Predictive Performance Comparison
4.3 Trading Performance Comparison
4.4 Ablation Study
4.5 Visualization of Proposed Loss Function
5 Conclusion
References
MOT: A Mixture of Actors Reinforcement Learning Method by Optimal Transport for Algorithmic Trading
1 Introduction
2 Problem Formulation
3 Methodology
3.1 Imitation Learning
3.2 Pretrain Module
3.3 Multiple Actors
3.4 Optimal Transport Regularization
4 Experiments
4.1 Dataset
4.2 Baselines, Evaluation Metrics and Hyperparameters
4.3 Experimental Results
4.4 Ablation Study
5 Related Work
6 Conclusion
References
Agent-Based Simulation of Decision-Making Under Uncertainty to Study Financial Precarity
1 Introduction
2 Background: Modeling Consumption
2.1 Capturing Uncertainty
3 The Framework: Introducing Real Constraints
3.1 Background: Modeling Ruin
3.2 Our New Model
4 Simulation Study: Precarity
4.1 Long Term Precarity
4.2 Factors Contributing to Precarity
5 Simulation Study: Interventions
6 Related Work
7 Conclusions
References
Information Retrieval and Search
Semantic Completion: Enhancing Image-Text Retrieval with Information Extraction and Compression
1 Introduction
2 Related Work
2.1 Dual-Stream Structure
2.2 Single-Stream Structure
3 Methodology
3.1 Overview
3.2 Information Extraction and Compression (IEC)
3.3 Training Tasks
4 Experiments
4.1 Experimental Settings
4.2 Comparison to Baseline
4.3 Ablation Study
5 Conclusion
References
Fast Edit Distance Prediction for All Pairs of Sequences in Very Large NGS Datasets
1 Introduction
2 Related Works
3 Basic Concepts
3.1 Preliminary
3.2 Key Ideas
3.3 Selecting the Reference Sequences
4 Predicting the Edit Distance
5 Edit Distance Prediction from Non-matching Sub-sequences
5.1 Differences of Distances with Respect to Pairs of Letters
5.2 Computational Complexity
6 Results and Discussions
6.1 Datasets
6.2 Selecting K-Mer Length
6.3 Comparison of the Proposed Strategies
6.4 Edit Distance Prediction
6.5 Hierarchical Clustering
7 Conclusion
References
MixCL: Mixed Contrastive Learning for Relation Extraction
1 Introduction
2 Mixed Contrastive Learning
2.1 Neural Relation Extraction Baseline Model
2.2 Entity Centralized Contrastive Learning for Relation Extraction
2.3 Mixed Contrastive Learning for Relation Extraction
3 Experiments
3.1 Datasets and Comparison Models
3.2 Main Results
4 Analysis
4.1 Metric of Quality of Negative Examples
4.2 Why Does MixCL Work?
5 Related Work
5.1 Neural Relation Extraction
5.2 Contrastive Learning
6 Conclusion
References
Decomposing Relational Triple Extraction with Large Language Models for Better Generalization on Unseen Data
1 Introduction
2 Methodology
2.1 Problem Formalization
2.2 Sub-task 1: Relation Extraction
2.3 Sub-task 2: Entity Extraction
2.4 Sub-task 3: Triple Filtering
3 Experiments
3.1 Datasets
3.2 Compared Methods and Evaluation Metrics
3.3 Overall Performance
3.4 Model Generalization Evaluation
3.5 Ablation Studies
4 Related Work
5 Conclusion
References
Multi-Query Person Search with Transformers
1 Introduction
2 Method
2.1 Multi-Query Decoder
2.2 Detection and Identification Losses
3 Experiments
3.1 Datasets
3.2 Evaluation Metrics
3.3 Implementation Details
3.4 Comparison to the State-of-the-Art Approaches
3.5 Model Analysis
4 Conclusion
References
BioReX: Biomarker Information Extraction Inspired by Aspect-Based Sentiment Analysis
1 Introduction
2 Related Work
3 Methodology
3.1 Biomarker Extraction (BE)
3.2 Result Association (RA)
4 Experiments
4.1 Experiment Setting
4.2 Results and Analysis
5 Conclusions
References
IR Embedding Fairness Inspection via Contrastive Learning and Human-AI Collaborative Intelligence
1 Introduction
2 Related Works
3 Methodology
3.1 Model Overview
3.2 GP Based Contrastive Learning
4 Experiments and Deployment
4.1 Implementation
4.2 Offline Evaluation
4.3 Online A/B Testing
5 Conclusion
References
SemPool: Simple, Robust, and Interpretable KG Pooling for Enhancing Language Models
1 Introduction
2 Related Work
3 Problem Statement and Preliminaries
4 Empirical Findings on Robustness
5 SemPool: Semantic Graph Pooling
5.1 KG Initialization
5.2 Pooling
5.3 KG Grounding
6 Experimental Setting
7 Results
7.1 Main Results
7.2 Ablation Studies and Analysis
8 Conclusions
References
Medical and Biological Data
Spatial Gene Expression Prediction Using Multi-Neighborhood Network with Reconstructing Attention
1 Introduction
2 Related Work
2.1 Gene Expression Prediction
2.2 Vision Transformer (ViT)
3 Method
3.1 Dual-Scale Attention
3.2 Reconstructing Attention
3.3 Transformer Block
4 Experiment
4.1 Datasets
4.2 Experiment Setup
5 Results and Discussion
5.1 Baseline Experiments for Multi-Neighborhood Network
5.2 Influence of Number of Neighborhoods
5.3 Influence of Attention Mechanisms
6 Conclusion
References
APFL: Active-Passive Forgery Localization for Medical Images
1 Introduction
2 Related Work
2.1 Active Forgery Localization
2.2 Passive Forgery Localization
3 Methodology
3.1 Active Fuzzy Localization with Reversible Watermarking
3.2 Passive Precise Localization with Lightweight KDU-Net
4 Experiments
4.1 Settings
4.2 Comparison Results
4.3 Robustness Evaluation
4.4 Hyperparameter Evaluation
4.5 Ablation Study
5 Conclusion
References
A Universal Non-parametric Approach for Improved Molecular Sequence Analysis
1 Introduction
2 Related Work
3 Proposed Approach
3.1 Compression Methods
3.2 Problem Formulation
3.3 Our Algorithm
3.4 Distance Matrix Symmetry
3.5 Kernel Matrix Computation
3.6 Experimental Setup
3.7 Justification of Employing the Kernel Matrix
4 Results and Discussion
5 Conclusion
References
Dynamic GNNs for Precise Seizure Detection and Classification from EEG Data
1 Introduction
2 Related Work
3 Methodology: NeuroGNN
3.1 Node Features Generation
3.2 Adjacency Matrix Generation
3.3 Prediction Using the Generated NeuroGraph
3.4 Pretraining
4 Experiments
4.1 Experimental Setup
4.2 Results
4.3 Ablation Study
4.4 Analysis of Graph Embeddings
4.5 Handling Scarce Training Data
5 Conclusion
References
A Novel Population Graph Neural Network Based on Functional Connectivity for Mental Disorders Detection
1 Introduction
2 Methods
2.1 Brain Connectomic Graph
2.2 Heterogeneous Population Graph
3 Dataset and Experimental Evaluation
3.1 Dataset and Experimental Setup
3.2 Cross-Validation
3.3 Ablation Experiment and Parameter Sensitivity Analysis
3.4 Interpretability Analysis
4 Conclusion
References
Weighted Chaos Game Representation for Molecular Sequence Classification
1 Introduction
2 Related Work
3 Proposed Approach
3.1 Vision DL Models
4 Experimental Evaluation
4.1 Baselines Models:
5 Results and Discussion
6 Conclusion
References
Robust Influence-Based Training Methods for Noisy Brain MRI
1 Introduction
2 Related Work
3 Preliminaries
3.1 Influence Function on Single Validation Point
3.2 Influence Function on Validation Group
4 Influence-Based Sample Reweighing
4.1 Framework
4.2 Calculating Sample Weights
5 Influence-Based Sample Perturbation
5.1 Framework
5.2 Selecting Influential Samples
5.3 Adding Influence-Based Perturbation
6 Experiments
6.1 Evaluation Setup
6.2 Proposed Methods and Baselines
6.3 Results
7 Conclusion
References
Co-ReaSON: EEG-based Onset Detection of Focal Epileptic Seizures with Multimodal Feature Representations
1 Introduction
2 Problem Definition
3 The Co-ReaSON Approach
3.1 Data Preprocessing
3.2 Feature Extraction
3.3 Co-ReaSON Predictive Model
4 Evaluation Setup
5 Evaluation Results
6 Related Work
7 Conclusion
References
A Data-Driven Approach for Building a Cardiovascular Disease Risk Prediction System
1 Introduction
2 The Proposed AutoML System
3 Experiment
3.1 The Dataset
3.2 Evaluation Metrics
3.3 Pipelines
3.4 Computation Cost
3.5 Comparison with Other AutoML Systems
4 Conclusion
A Appendix
References
TFAugment: A Key Frequency-Driven Data Augmentation Method for Human Activity Recognition
1 Introduction
2 Related Work
2.1 Data Augmentation Approaches
2.2 Deep Learning Approaches
3 Methodology
3.1 Time-Frequency Transformation
3.2 Time-Frequency Augmentation
3.3 TFAugment Based HAR Method
4 Experiments
4.1 Datasets
4.2 Methods for Comparison
4.3 Experimental Results
4.4 Discussion
5 Conclusion
References
Dfp-Unet: A Biomedical Image Segmentation Method Based on Deformable Convolution and Feature Pyramid
1 Introduction
2 Dfp-Unet
2.1 Feature Pyramid Network
2.2 Deformable Convolution
2.3 Structure of Dfp-Unet
3 Experimental Data and Process
3.1 Description of Datasets
3.2 Experimental Environment and Process
4 Experimental Results and Analysis
5 Discussion
References
ACHIM: Adaptive Clinical Latent Hierarchy Construction and Information Fusion Model for Healthcare Knowledge Representation
1 Introduction
2 Related Work
3 Methods
3.1 Time Series Extractor
3.2 Latent Hierarchy Refactorer
3.3 Information Aggregator
3.4 Healthcare Prediction
4 Experiment
4.1 Experimental Results
4.2 Ablation Experiments
4.3 Qualitative Visualization of Patient Representation
5 Conclusion
References
Using Multimodal Data to Improve Precision of Inpatient Event Timelines
1 Introduction
2 Methods
2.1 Annotation Tool
2.2 Annotation Process
2.3 Statistical Methods
2.4 Multimodal Learning
3 Results
3.1 Comparison Across Different Modalities
3.2 Comparison with I2b2 Dataset (2012)
3.3 Multimodal Learning
4 Discussion and Conclusions
References
Adversarial-Robust Transfer Learning for Medical Imaging via Domain Assimilation
1 Introduction
2 Related Work
3 Preliminaries
4 Method
5 Experiments
5.1 Dataset
5.2 Setup
5.3 Evaluation
5.4 Results
6 Conclusion
References
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken