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

Preface
Organization
Contents – Part IV
Theory and Algorithms I
Knowledge Transfer from Situation Evaluation to Multi-agent Reinforcement Learning
	1 Introduction
	2 Related Works
	3 Background
	4 Methods
		4.1 Situation Evaluation Task
		4.2 Construction of Guiding Dense Reward
		4.3 Transfer of Situation Comprehension Network
	5 Experiments
		5.1 Performance on Google Research Football
		5.2 Parametric Sensitivity of GDR
		5.3 Transfer Layers Study
	6 Conclusion
	References
Sequential Three-Way Rules Class-Overlap Under-Sampling Based on Fuzzy Hierarchical Subspace for Imbalanced Data
	1 Introduction
	2 Related Work
	3 Sequential Three-Way Rules Class-Overlap Under-Sampling Based on Fuzzy Hierarchical Subspace (S3RCU)
	4 Experimental Results and Discussion
		4.1 Datasets
		4.2 Experimental Settings
		4.3 Experimental Results
		4.4 Global Analysis of Results
	5 Conclusion and Future Work
	References
Two-Stage Multilayer Perceptron Hawkes Process
	1 Introduction
	2 Related Work
	3 The Proposed TMPHP Model
		3.1 Two-Stage Multilayer Perceptron Hawkes Process
		3.2 Conditional Intensity Function
		3.3 Loss Function and the Logarithmic Likelihood Function
	4 Experiments
		4.1 Datasets
	5 Experimental Results and Comparison
	6 Conclusions and Future Works
	References
The Context Hierarchical Contrastive Learning for Time Series in Frequency Domain
	1 Introduction
	2 The Proposed CHCL-TSFD Model
		2.1 Problem Definition for Time Series Representation Learning
		2.2 The Proposed CHCL-TSFD Framework for Time Series Representation Learning
	3 The Proposed Context Hierarchical Contrasting Learning
	4 Experimental Results
		4.1 Time Series Classification Problem
		4.2 Time Series Forecasting
	5 Conclusions and Future Works
	References
Hawkes Process via Graph Contrastive Discriminant Representation Learning and Transformer Capturing Long-Term Dependencies
	1 Introduction
	2 Related Work
	3 Model
		3.1 Graph Contrastive Discriminant Representation Learning for Transformer Hawkes Process
		3.2 Conditional Intensity Function
		3.3 Loss Function
	4 Experiments
		4.1 Datasets
		4.2 Experimental Comparative Results
	5 Conclusions and Future Work
	References
A Temporal Consistency Enhancement Algorithm Based on Pixel Flicker Correction
	1 Introduction
	2 Related Work
		2.1 Task-Dependent Algorithms
		2.2 Post-processing Algorithms
	3 Methodology
		3.1 Temporal Stability Module
		3.2 Hybrid Loss Function
	4 Experiments
		4.1 Datasets
		4.2 Model Training and Inference
		4.3 Evaluation Metrics
		4.4 Experimental Results and Analysis
		4.5 Ablation Experiments
	5 Conclusion
	References
Data Representation and Clustering with Double Low-Rank Constraints
	1 Introduction
	2 Data Clutering with Double Low-Rank Constraints
		2.1 The UV-LRR Model
		2.2 The Robust UV-LRR Model
		2.3 Complexity Analysis
	3 Experiments
		3.1 Experimental Settings
		3.2 Clustering Quality Evaluation
		3.3 Robustness to Noise
		3.4 Convergence Analysis
		3.5 Parameter Sensitivity Analysis
	4 Conclusion
	References
RoMA: A Method for Neural Network Robustness Measurement and Assessment
	1 Introduction
	2 Background
	3 The Proposed Method
		3.1 Probabilistic Robustness
		3.2 Sampling Method and the Normal Distribution
		3.3 The Box-Cox Transformation
		3.4 The RoMA Certification Algorithm
	4 Evaluation
	5 Summary and Discussion
	References
Independent Relationship Detection for Real-Time Scene Graph Generation
	1 Introduction
	2 Related Work
	3 Method
		3.1 Independent Relationship Detection
		3.2 Scene Graph Generation
	4 Experiment
		4.1 Dataset, Model and Metrics
		4.2 Results and Discussion
	5 Conclusion
	References
A Multi-label Feature Selection Method Based on Feature Graph with Ridge Regression and Eigenvector Centrality
	1 Introduction
	2 The Proposed Method
		2.1 Exploring Feature Label Correlation via Ridge Regression
		2.2 Feature Ranking with Eigenvector Centrality
	3 Experiments
		3.1 Experimental Data Sets
		3.2 Experimental Settings and Compared Methods
		3.3 Results and Discussion
	4 Conclusion
	References
O3GPT: A Guidance-Oriented Periodic Testing Framework with Online Learning, Online Testing, and Online Feedback
	1 Introduction
	2 Related Works
	3 Methods and Implementation
		3.1 Problem Definition and Notations
		3.2 The Overall Architecture of the Proposed Model
	4 Experiment Settings
		4.1 Datasets Description
		4.2 Baselines
		4.3 Evaluation Metrics
	5 Experimental Results and Analysis
	6 Conclusions
	References
AFFSRN: Attention-Based Feature Fusion Super-Resolution Network
	1 Introduction
	2 Related Work
	3 Method
		3.1 Overview
		3.2 Deep Feature Fusion Group
		3.3 Self-attention Feature Distillation Block
	4 Experiments
		4.1 Tranining Detail
		4.2 Comparisons with State-of-the-Arts Methods
		4.3 Ablation Studies
		4.4 Model Complexity Analysis
	5 Conclusion
	References
Temporal-Sequential Learning with Columnar-Structured Spiking Neural Networks
	1 Introduction
	2 Model Description
		2.1 Network Architecture
		2.2 Spiking Neuron Model Inspired by Pyramidal Cell
		2.3 Learning
		2.4 Retrieval
	3 Experiment
		3.1 Goal-Based Retrieval
		3.2 Context-Based Retrieval
		3.3 Parameter Influence
	4 Conclusion
	References
Graph Attention Transformer Network for Robust Visual Tracking
	1 Introduction
	2 Related Work
		2.1 Visual Tracking
		2.2 Attention for Tracking
	3 Proposed Method
		3.1 Overview
		3.2 Transformer-Based Feature Extraction
		3.3 Adaptive Graph Attention Module
	4 Experiments
		4.1 Implementation Details
		4.2 Ablation Study
		4.3 Comparisons with State-of-the-Art Trackers
	5 Conclusion
	References
GCL-KGE: Graph Contrastive Learning for Knowledge Graph Embedding
	1 Introduction
	2 Related Works
	3 Proposed Model
		3.1 Overview
		3.2 Encoder
		3.3 Decoder and Score
		3.4 Contrastive Learning
		3.5 Training Objective
		3.6 Theoretical Analyses
	4 Experiments
		4.1 Experimental Setup
		4.2 Main Results
		4.3 Ablation Experiments
	5 Conclusion
	References
Towards a Unified Benchmark for Reinforcement Learning in Sparse Reward Environments
	1 Related Work
	2 Building the Benchmark
		2.1 Quantify Sparse Reward Environments with Sparsity
		2.2 Sparsity Inspired RL Algorithm Categorization
		2.3 ARTS: A New Evaluation Metric via Sparsity
		2.4 Benchmark Construction
	3 Experimental Evaluations
		3.1 Measuring the Sparsity of Environments
		3.2 Algorithm Evaluations
	4 Conclusive Remarks
	References
Effect of Logistic Activation Function and Multiplicative Input Noise on DNN-kWTA Model
	1 Introduction
	2 Basic DNN-kWTA
	3 Logistic DNN-kWTA with Input Noise
		3.1 DNN-kKWTA Under Imperfection
		3.2 Equivalent Model
	4 Non-gaussian Multiplicative Input Noise
	5 Simulation Results
		5.1 Effectiveness of Theorem 4
		5.2 Effectiveness of Theorem 5
		5.3 Effectiveness of Theorem 7
	6 Conclusion
	References
A High-Speed SSVEP-Based Speller Using Continuous Spelling Method
	1 Introduction
	2 Materials
		2.1 Participants
		2.2 Visual Stimulus Presentation
		2.3 Experiments
	3 Methods
		3.1 Data Processing
		3.2 Template Reconstruction CCA
		3.3 Performance Evaluation
	4 Results
		4.1 Offline BCI Performance
		4.2 Online BCI Performance
	5 Discussions
	6 Conclusion
	References
AAT: Non-local Networks for Sim-to-Real Adversarial Augmentation Transfer
	1 Introduction
	2 Method
		2.1 Conditional Domain Adversarial Networks  ch19long2017conditional
		2.2 Semantic Data Augmentation Loss
		2.3 Non-local Attention
		2.4 Overall Networks
	3 Experiments
		3.1 Data Preparation
		3.2 Implementation Details
		3.3 Results
		3.4 Analysis
	4 Conclusion
	References
Aggregating Intra-class and Inter-class Information for Multi-label Text Classification
	1 Introduction
	2 Related Work
	3 Methods
		3.1 AIIF Models
		3.2 AIIF Training
		3.3 Label Graph Construction
	4 Experiments
		4.1 Datasets and Evaluations
		4.2 Baselines
		4.3 Implementation Details
		4.4 Main Results
		4.5 Ablation Study
		4.6 Performance on the Tail Labels
		4.7 Case Study
	5 Conclusion
	References
Fast Estimation of Multidimensional Regression Functions by the Parzen Kernel-Based Method
	1 Introduction
	2 Efficient Nonparametric Density Estimation Algorithms
	3 A Synopsis of the Kernel Methods in Regression Estimation
	4 The Novel Algorithm Basics
	5 The Scaling Factor an Selection and Its Impact on the Computational Speed
	6 Practical Applications of the Algorithm
	7 Conclusions
	References
ReGAE: Graph Autoencoder Based on Recursive Neural Networks
	1 Introduction
	2 Related Work
	3 Method
		3.1 Order of Vertices and Adjacency Matrix Patches
		3.2 Computational Complexity
		3.3 Variational Graph Autoencoder
		3.4 Further Extensions
	4 Experimental Study
		4.1 Datasets
		4.2 Graph Autoencoding
	5 Conclusions
	References
Efficient Uncertainty Quantification for Under-Constraint Prediction Following Learning Using MCMC
	1 Introduction
	2 Data and Model
		2.1 Learning Parameters of the Logistic Regression Model
	3 Generating Training Data Sets at Parameter Summaries
		3.1 Learning  Under the Constraint Given Dmean
		3.2 DNN Results
		3.3 Learning  Under Constraint, with MCMC, Using Dright
	4 Pipe-Lined Architecture for Efficient Uncertainty Percolation
	5 Discussion and Conclusions
	References
SMART: A Robustness Evaluation Framework for Neural Networks
	1 Introduction
	2 Related Work
		2.1 Adversarial Examples
		2.2 Robustness Evaluation
		2.3 Separability Index
	3 Method
		3.1 Model\'s Robustness and Data Separability
		3.2 The Separability Index MDSI
		3.3 The Robustness Evaluation Framework SMART
	4 Experiments
		4.1 The Validity of SMART
		4.2 SMART and Mainstream Robustness Metrics
	5 Conclusion
	References
Time-aware Quaternion Convolutional Network for Temporal Knowledge Graph Reasoning
	1 Introduction
	2 Related Works
	3 Preliminary
	4 Overview of T-QGCN
		4.1 Encoding Module
		4.2 Decoding Module
		4.3 Parameter Learning
	5 Experiment
		5.1 Experimental Setup
		5.2 Experimental Results
		5.3 Ablation Studies
	6 Conclusion
	References
SumBART - An Improved BART Model for Abstractive Text Summarization
	1 Introduction
	2 Literature Survey
	3 Methodology
		3.1 Additional Embedding Using Information from Knowledge Graph
		3.2 Keyword Extractor
		3.3 CNN Embedding
	4 Implementation and Results
	5 Conclusions
	References
Saliency-Guided Learned Image Compression for Object Detection
	1 Introduction
	2 Proposed Method
		2.1 The Overall Framework
		2.2 Saliency Analysis for Machine Vision
		2.3 Saliency-Guided Bitrate Allocation
		2.4 Coding Optimization for Machine Vision
	3 Experiment and Results
		3.1 Implementation Details
		3.2 Effectiveness of Saliency-Guided Bitrate Allocation
		3.3 The Overall Performance Evaluation
	4 Conclusion
	References
Multi-label Learning with Data Self-augmentation*-12pt
	1 Introduction
	2 Related Works
	3 Methods
		3.1 Problem Formulation
		3.2 Data Augmentation for Multi-label Learning
		3.3 Graph Laplacian Regularization
		3.4 Optimization
	4 Experiments
		4.1 Datasets and Experiment Settings
		4.2 Experimental Results
		4.3 Parameter Sensitivity Analysis
	5 Conclusion
	References
MnRec: A News Recommendation Fusion Model Combining Multi-granularity Information
	1 Introduction
	2 Our Approach
		2.1 Word Embedding
		2.2 Cross-Attention
		2.3 Granular Network
		2.4 Representation Layer and Representation Attention
		2.5 RTCN Hierarchical Interest
		2.6 Click Predictor
	3 Experiments
		3.1 Experiment Setup
		3.2 Performance Evaluation
		3.3 Ablation Studies
		3.4 Parameter Analysis
		3.5 Conclusion
	References
Infinite Label Selection Method for Mutil-label Classification
	1 Introduction
	2 Preliminaries
	3 The Proposed Method
		3.1 Method for Label Subset Selection
		3.2 Recovery of the Selected Label Subset
	4 Experiments
		4.1 Experimental Setup
		4.2 Experimental Settings
		4.3 Results and Discussion
	5 Conclusion
	References
Simultaneous Perturbation Method for Multi-task Weight Optimization in One-Shot Meta-learning
	1 Introduction
	2 Related Works
	3 Multi-task Meta-learning Modification
		3.1 One-Shot Learning Problem Definition
		3.2 Multi-Task Meta-Learning Loss Function
		3.3 Multi-task Weights Optimization
		3.4 SPSA for Tracking Method
	4 Experiments
	5 Ablation Study
	6 Conclusion
	References
Searching for Textual Adversarial Examples with Learned Strategy
	1 Introduction
	2 Related Work
	3 Methodology
		3.1 Problem Definition
		3.2 Synonym Selection Network
		3.3 Beam Search
		3.4 MCTS
	4 Experiments
		4.1 Datasets and Victim Models
		4.2 Experiment Setup
		4.3 Evaluation Metrics
		4.4 Main Results
		4.5 Decomposition Analysis
		4.6 Transferability
	5 Conclusion
	References
Multivariate Time Series Retrieval with Binary Coding from Transformer
	1 Introduction
	2 Related Work
	3 Multivariate Time-Series Retrieval Network
		3.1 Problem Statement
		3.2 Transformer-Based Encoder
		3.3 Transformer-Based Decoder
		3.4 Adversarial Loss
		3.5 Hashing Loss
		3.6 Binary Codes Embedding
		3.7 Objective and Training Procedure
	4 Experiments
		4.1 Datasets
		4.2 Parameters Setting and Evaluation Metrics
		4.3 Results
	5 Conclusion
	References
Learning TSP Combinatorial Search and Optimization with Heuristic Search*-12pt
	1 Introduction
	2 Related Works
	3 The Model Architecture
		3.1 Model Designing
		3.2 Model Training with Reinforcement Learning
	4 Experiments
		4.1 Datasets and Experimental Details
		4.2 Results and Analysis
	5 Conclusion
	References
A Joint Learning Model for Open Set Recognition with Post-processing
	1 Introduction
	2 Preliminaries
	3 Proposed Method
		3.1 Joint Learning with Self-supervision
		3.2 Learning Dynamic Boundary
		3.3 Post-processing Penalty Mechanism
		3.4 Learning the Open Set Network
	4 Experiments
		4.1 Implementation Details
		4.2 Open Set Recognition
		4.3 Ablation Study
	5 Conclusion
	References
Cross-Layer Fusion for Feature Distillation*-12pt
	1 Introduction
	2 Related Work
		2.1 Knowledge Distillation
		2.2 Feature-Map Distillation
		2.3 Feature Fusion Method
	3 Method
		3.1 Notations
		3.2 Logit-Based Knowledge Distillation
		3.3 Teacher-Student Feature Fusion
		3.4 Fusion Module and Dynamic Feature Fusion Strategy
	4 Experiments
		4.1 Experimental Setup
		4.2 Experiments of CIFAR-10 Dataset
		4.3 Experiments of CIFAR-100 Dataset
		4.4 Ablation Study
	5 Conclusion
	References
MCHPT: A Weakly Supervise Based Merchant Pre-trained Model*-12pt
	1 Introduction
	2 Related Work
		2.1 Domain-specific PTMs
		2.2 Weakly Supervised PTMs
	3 MCHPT Model
		3.1 Weakly Supervised Dataset Construction
		3.2 Pre-training Tasks
	4 Experiments
		4.1 Merchants Related Tasks
		4.2 Experimental Setup
		4.3 Experiment Results
		4.4 Ablation Studies
	5 Conclusions and Future Work
	References
Progressive Latent Replay for Efficient Generative Rehearsal
	1 Introduction
	2 Related Work
	3 Method
		3.1 Progressive Latent Replay
		3.2 Computational Efficiency of the Method
	4 Experiments
		4.1 Experimental Setup
		4.2 Model Architecture
		4.3 Metrics
		4.4 Pretraining Impact on Internal Replay
		4.5 Progressive Latent Replay with Different Strategies
	5 Conclusion
	References
Generalization Bounds for Set-to-Set Matching with Negative Sampling
	1 Introduction
	2 Preliminaries
		2.1 Set-to-Set Matching with Negative Sampling
	3 Margin Bound for Set-to-Set Matching
	4 RKHS Bound for Set-to-Set Matching
	5 Conclusion and Discussion
	References
ADA: An Attention-Based Data Augmentation Approach to Handle Imbalanced Textual Datasets
	1 Introduction
	2 Related Works
	3 Proposed Data Augmentation Approach
		3.1 Vector Similarity-Based Keywords Extraction
		3.2 Keywords-Based Labeled Dataset Creation
		3.3 Attention-Based Significant Words Extraction
		3.4 Language Model-Based Transformation of Review Documents
		3.5 Oversampling Minority Class Dataset
	4 Experimental Setup and Results
		4.1 Datasets
		4.2 Data Preprocessing
		4.3 Classifier Architecture and Training Details
		4.4 Evaluation Metrics
		4.5 Comparison Approaches
		4.6 Evaluation Results and Comparative Analysis
	5 Conclusion and Future Work
	References
Countering the Anti-detection Adversarial Attacks
	1 Introduction
	2 Background
	3 The Proposed Method
		3.1 The Residuals Used for the Proposed Feature
		3.2 The Proposed Feature
	4 Experimental Results
		4.1 Experiment Settings
		4.2 Comparative Results
		4.3 Discussions of Generalization Ability and Security
	5 Conclusion
	References
Evolving Temporal Knowledge Graphs by Iterative Spatio-Temporal Walks*-12pt
	1 Introduction
	2 Related Work
		2.1 Static KG Representation Learning
		2.2 Temporal KG Representation Learning
	3 Our Model
		3.1 Model Overview
		3.2 Spatio-temporal Walking
		3.3 TRGAT
	4 Experiments
		4.1 Experimental Setup
		4.2 Results on TKG Reasoning
		4.3 Ablation Study
	5 Conclusions
	References
Improving Knowledge Graph Embedding Using Dynamic Aggregation of Neighbor Information*-12pt
	1 Introduction
	2 Related Work
	3 Proposed Model
		3.1 Neighborhood Information Transformation
		3.2 Dynamic Aggregation of Neighbor Information
	4 Experiments
		4.1 Experimental Settings
		4.2 Experimental Results
	5 Conclusion
	References
Generative Generalized Zero-Shot Learning Based on Auxiliary-Features
	1 Introduction
	2 Related Work
	3 Our Proposed Method
		3.1 Overall Procedure of Our Method
		3.2 Learning the Instructive-Features
		3.3 Auxiliary-features via GAN(Af-GAN)
	4 Experiments
		4.1 On ZSL
		4.2 On GZSL
		4.3 Ablation Study and Experiment Analysis
	5 Conclusion
	References
Learning Stable Representations with Progressive Autoencoder (PAE)
	1 Introduction
	2 Framework of Progressive Autoencoder
		2.1 Encoder Network
		2.2 Progressive Patching Decoder
		2.3 Training Method
	3 Experiments
		3.1 Non-linear Toy Dataset
		3.2 Experiment Settings
		3.3 Results of Toy Dataset
		3.4 Comparison Between Beta-PAE and Beta-VAE
		3.5 Experiments on MNIST
	4 Discussion and Conclusion
	References
Effect of Image Down-sampling on Detection of Adversarial Examples
	1 Introduction
	2 Motivation Experiment
	3 Further Investigation
		3.1 Down-Sampling Algorithm
		3.2 Adversarial Attacks
		3.3 Detection Results of ESRM for Different Interpolation Kernels
		3.4 Detection Results of FS for Different Interpolation Kernels
		3.5 Discussion
	4 Conclusion
	References
Boosting the Robustness of Neural Networks with M-PGD
	1 Introduction
	2 Backgrounds
		2.1 Terminology and Notation
		2.2 Attack Algorithms
		2.3 Defensive Algorithms
	3 M-PGD Adversarial Training Algorithm
		3.1 M-PGD Attack
		3.2 Adversarial Training
	4 Experiments
		4.1 Parameters Selection
		4.2 Models Comparison
		4.3 Adversarial Loss Value
	5 Conclusions
	References
StatMix: Data Augmentation Method that Relies on Image Statistics in Federated Learning
	1 Introductions
		1.1 Motivation
		1.2 Contribution
	2 Related Work
	3 Proposed Approach
	4 Experimental Setup
	5 Experimental Results and Analysis
		5.1 Ablation Study
	6 Concluding Remarks
	References
Classification by Components Including Chow\'s Reject Option
	1 Introduction
	2 The Original CbC
	3 Reject Options for a Probabilistic Classifier
	4 Reject Options for CbC Networks
	5 Derivation and Comparison of Gradients
	6 Experiments and Simulations
	7 Conclusion
	A  Proof of Optimality of Chow\'s Threshold
	References
Community Discovery Algorithm Based on Improved Deep Sparse Autoencoder
	1 Introduction
	2 Related Work
		2.1 Community Discovery
		2.2 Deep Sparse Autoencoder
	3 Method
		3.1 Enhance Node Similarity
		3.2 Feature Extraction
	4 Experimental Results and Analysis
		4.1 Experiment Preparation
		4.2 Evaluation Criteria and Comparison Algorithms
		4.3 Experimental Results
	5 Conclusions and Future Work
	References
Fairly Constricted Multi-objective Particle Swarm Optimization
	1 Introduction
	2 Motivations
	3 Finding a Fairly Constricted Algorithm
	4 Results
		4.1 Assessment with Quality Indicators
	5 Discussion, Conclusion and Future Works
	References
Argument Classification with BERT Plus Contextual, Structural and Syntactic Features as Text
	1 Introduction
	2 Related Works
	3 Datasets
	4 Model
		4.1 BERT
		4.2 Features
		4.3 Combined Features as Text
	5 Results and Analysis
		5.1 Experimental Setting
		5.2 Task Results
		5.3 Analysis
	6 Conclusion
	References
Variance Reduction for Deep Q-Learning Using Stochastic Recursive Gradient
	1 Introduction
	2 Related Work
		2.1 SGD for Deep Q-Learning
		2.2 Variance Reduced Deep Q-Learning
	3 Our Approach: SRG-DQN
		3.1 Problem Analysis
		3.2 Recursive Gradient Deep Q-Learning
		3.3 Theoretical Analysis
	4 Experiments
		4.1 Experimental Settings
		4.2 Experimental Results
		4.3 Experimental Analysis
	5 Conclusion
	References
Optimizing Knowledge Distillation via Shallow Texture Knowledge Transfer
	1 Introduction
	2 Related Work
		2.1 Knowledge Distillation
		2.2 Gradient Local Binary Pattern
		2.3 Attention Mechanism
	3 Method
		3.1 Preliminary Research and Notation
		3.2 Shallow Texture Knowledge Distillation
		3.3 Texture Attention Module
		3.4 Implementation Details
	4 Experiments
		4.1 Experiments on the CIFAR Dataset
		4.2 Ablation Experiments
	5 Conclusion
	References
Unsupervised Domain Adaptation Supplemented with Generated Images
	1 Introduction
	2 Related Work
	3 Problem Statement
	4 Method
		4.1 Generative Module
		4.2 UDA Algorithm Module
	5 Experiments
		5.1 Implementation
		5.2 Results
	6 Conclusion and Future Work
	References
MAR2MIX: A Novel Model for Dynamic Problem in Multi-agent Reinforcement Learning
	1 Introduction
	2 Background
		2.1 Problem Formulation
		2.2 Deep Deterministic Policy Gradient
	3 Method
		3.1 RNN Augmented SAC
		3.2 Dynamic Mask Technology
		3.3 Recurrent Residual MIX Network
	4 Experiments
		4.1 Environmental Setup
		4.2 Parameter Setting
		4.3 Experimental Results
		4.4 Ablation Study
	5 Conclusion
	References
Adversarial Training with Knowledge Distillation Considering Intermediate Representations in CNNs
	1 Introduction
	2 Experimental Analysis in CNNs with Adversarial Training
		2.1 Adversarial Training
		2.2 Visualization of Intermediate Representations in CNNs
	3 Proposed Method: Adversarial Training with Knowledge Distillation
		3.1 Knowledge Distillation
		3.2 Adversarial Training with Knowledge Distillation
	4 Experimental Evaluation
		4.1 Experimental Setup
		4.2 Classification Accuracy
		4.3 Visualization of Intermediate Representations
	5 Conclusions
	References
Deep Contrastive Multi-view Subspace Clustering
	1 Introduction
	2 Related Works
		2.1 Subspace Clustering
		2.2 Multi-view Subspace Clustering
		2.3 Contrastive Learning
	3 Proposed Method
		3.1 The Proposed DCMSC
		3.2 Optimization
	4 Experiment
		4.1 Experimental Settings
		4.2 Experimental Results
		4.3 Visualization
		4.4 Ablation Studies
	5 Conclusion
	References
Author Index