Neural Computing for Advanced Applications: Third International Conference, NCAA 2022, Jinan, China, July 8–10, 2022, Proceedings, Part I (Communications in Computer and Information Science, 1637)

The two-volume Proceedings set CCIS 1637 and 1638 constitutes the refereed proceedings of the Third International Conference on Neural Computing for Advanced Applications, NCAA 2022, held in Jinan, China, during July 8–10, 2022. The 77 papers included in these proceedings were carefully reviewed and selected from 205 submissions. These papers were categorized into 10 technical tracks, i.e., neural network theory, and cognitive sciences, machine learning, data mining, data security & privacy protection, and data-driven applications, computational intelligence, nature-inspired optimizers, and their engineering applications, cloud/edge/fog computing, the Internet of Things/Vehicles (IoT/IoV), and their system optimization, control systems, network synchronization, system integration, and industrial artificial intelligence, fuzzy logic, neuro-fuzzy systems, decision making, and their applications in management sciences, computer vision, image processing, and theirindustrial applications, natural language processing, machine translation, knowledge graphs, and their applications, Neural computing-based fault diagnosis, fault forecasting, prognostic management, and system modeling, and Spreading dynamics, forecasting, and other intelligent techniques against coronavirus disease (COVID-19). Preface Organization Contents – Part I Contents – Part II TE-BiLSTM: Improved Transformer and BiLSTM on Fraudulent Phone Text Recognition 1 Introduction 2 Related Work 3 Model Architecture 3.1 TE - BiLSTM Model 3.2 Improved Transformer 3.3 BiLSTM Network Layer 4 Experiment and Analysis 4.1 Dataset 4.2 Evaluation Criteria 4.3 Parameter Settings 4.4 Experimental Results and Discussion 5 Conclusions References Cross Elitist Learning Multifactorial Evolutionary Algorithm 1 Introduction 2 Preliminaries 2.1 Multifactorial Evolutionary Algorithm 2.2 Nelder-Mead Algorithm 2.3 Opposition Learning 3 Proposed Method 3.1 Cross Elitist Learning Strategy 3.2 Knowledge Transfer Based on Nelder-Mead Method 3.3 Opposition Learning 3.4 The Framework of CEL-MFEA 3.5 The Computional Time Analysis 4 Comparative Studies of Experiments 4.1 General Experimental Setting and Parameter Setting 4.2 Comparative Experiments 5 Conclusions References Heterogeneous Adaptive Denoising Networks for Recommendation 1 Introduction 2 Related Work 3 Model Architecture 3.1 Input and Embedding Module 3.2 Recommendation and ADT Module 4 Experiments 4.1 Evaluation Datasets 4.2 Overall Performance 4.3 Ablation Experiments 4.4 Detailed Analysis of HGCRD 4.5 Limitation 5 Conclusion and Future Work References Formation Control Optimization via Leader Selection for Rotor Unmanned Aerial Vehicles 1 Introduction 2 System Model 3 Leader Selection Algorithm 4 Simulation Results 5 Conclusion References Dynamic Monitoring Method Based on Comparative Study of Power and Environmental Protection Indicators 1 Introduction 2 Short-term Emissions Forecasting Process 2.1 Building a Short-Term Electricity Load Forecasting Model 2.2 Constructing a Similarity Mapping Model for Electricity Consumption and Environmental Protection 2.3 Construction of Enterprise Pollution Prediction Model 3 Case Analysis 3.1 Sample Data Preprocessing 3.2 Data Spatial Distribution Visualization 3.3 Comparative Analysis of Results 4 Conclusion References Combustion State Recognition Method in Municipal Solid Waste Incineration Processes Based on Improved Deep Forest 1 Introduction 2 Processes Description and Modeling Strategy 3 Algorithm Implementation 3.1 Image Preprocessing Module 3.2 Improved Deep Forest Module 4 Experimental Verification 4.1 Description of Data 4.2 Results of Experiment 5 Conclusion References RPCA-Induced Graph Tensor Learning for Incomplete Multi-view Inferring and Clustering 1 Introduction 2 Related Work 2.1 Notation Summary 2.2 Adaptive Neighbors Graph Learning (ANGL) 2.3 Robust Principle Component Analysis (RPCA) 3 Proposed Method 3.1 Model of RPCA-IGTL 3.2 Optimization 4 Experiment 5 Conclusion References TRUST-TECH Assisted GA-SVM Ensembles and Its Applications 1 Introduction 2 Preliminaries 2.1 Support Vector Machine 2.2 Genetic Algorithms 2.3 GA-Combined Clustering 2.4 TRUST-TECH Technology 2.5 Ensemble of SVMs 3 Algorithm Procedures 3.1 SVM Parameters 3.2 GA-SVM for Parameter Optimization 3.3 Improvement on GA-SVM Using Clustering Method 3.4 Ensemble on SVM Using TRUST-TECH 4 Application of Algorithm in Regression 4.1 Problem Definition 4.2 Multiple Models Generated by GA-SVM 5 Conclusion and Future Work References An Early Prediction and Label Smoothing Alignment Strategy for User Intent Classification of Medical Queries 1 Introduction 2 Related Work 3 The EP-LSA Strategy 3.1 User Intention Feature Extraction 3.2 User Intent Alignment Based on Early Prediction 3.3 Supervised Signals by Label Smoothing 4 Experiments and Results 4.1 Datasets 4.2 Evaluation Metrics 4.3 Parameter Settings 4.4 Baseline Methods 4.5 The Results 5 Conclusions References An Improved Partition Filter Network for Entity-Relation Joint Extraction 1 Introduction 2 Related Work 3 Methods 3.1 The Dual-Joint-Input-PFN Strategy 3.2 The Dual-Joint-Input-PFN-Decoder Model 4 Experiment 4.1 Dataset 4.2 Evaluation Metrics 4.3 Baselines 4.4 The Results 5 Conclusion References Fast Dynamic Response Based on Active Disturbance Rejection Control of Dual Active Bridge DC-DC Converter 1 Introduction 2 Circuit Schematic of DAB 3 Output Steady State Current of the Converter 4 LADRC Model 5 Experimental Results 6 Conclusion References Adaptive Fuzzy Distributed Formation Tracking for Second-order Nonlinear Multi-agent Systems with Prescribed Performance 1 Introduction 2 Problem Formulation and Preliminaries 2.1 Algebraic Graph Theory 2.2 Useful Lemmas 2.3 The Dynamic Models of Multi-agent Systems 2.4 Control Objective 3 Main Results 4 Numerical Simulation 5 Conclusions References Path Planning for Mobile Robots Based on Improved A* Algorithm 1 Introduction 2 Related Work 2.1 A* Algorithm 2.2 A*-Connect 3 Four-Way A* Algorithm 3.1 Improvement I 3.2 Improvement II 3.3 Optimize A*-Connect Algorithm 4 Experimental Verification 4.1 Simulation Analysis 4.2 Experimental Verification 5 Conclusion References ML-TFN: Multi Layers Tensor Fusion Network for Affective Video Content Analysis 1 Introduction 2 Related Work 3 Approach 3.1 Modality Embedding Subnetworks 3.2 Tensor Fusion Network 3.3 Multi-layer Fusion 4 Experiments 4.1 Datasets 4.2 Implementation Details 4.3 Experimental Results 5 Conclusions References A Dominance-Based Many-Objective Artificial Bee Colony Algorithm 1 Introduction 2 Artificial Bee Colony Algorithm 3 Dominance-Based Many-Objective ABC (DMaOABC) 3.1 Framework of DMaOABC 3.2 New Fitness Function and Mating Selection Based on Secondary Dominance Criterion 3.3 Improved ABC 4 Experimental Study 5 Conclusion References Container Lead Seal Detection Based on Nano-CenterNet 1 Introduction 2 Container Lead Seal Detection Based on CenterNet 2.1 Image Analysis of Detection Target 2.2 Container Lead Seal Detection Based on CenterNet 3 Nano-CenterNet 3.1 ShuffleNetV2 3.2 GhostPAN 3.3 Context Detection Module and CBAM Module 3.4 GHM Loss Function 3.5 Data Augment 4 Experiments and Analysis 5 Conclusion References Backstepping Control of Air-Handling Unit for Indoor Temperature Regulation 1 Introduction 2 State Space Representation of the Test Room 3 Control Design of the AHU System 3.1 Backstepping Controller Design 3.2 Main Results 4 Simulation Results and Discussions 5 Conclusions References Laplacain Pair-Weight Vector Projection with Adaptive Neighbor Graph for Semi-supervised Learning 1 Introduction 2 Proposed Method 2.1 Formulations of ANG-LapPVP 2.2 Optimization of ANG-LapPVP 2.3 Strategy of Classification 2.4 Computational Complexity 3 Experiments 3.1 Experiments on Artificial Dataset 3.2 Experiments on Benchmark Datasets 4 Conclusion References A Collaborators Recommendation Method Based on Multi-feature Fusion 1 Introduction 2 Related Works 3 Collaborators Recommendation Method Based on Multi-feature Fusion 3.1 Overview of Our Framework 3.2 Content Features Extraction 3.3 Meta-path Features Extraction 3.4 Reconstruction of Co-authorship Network 3.5 Representation of Scholar Based on Reconstructed Network 4 Experiment 4.1 Dataset 4.2 Evaluation Criteria 4.3 Results and Comparison 4.4 Parametric Study 5 Conclusion References Design of Online ESN-ADP for Dissolved Oxygen Control in WWTP 1 Introduction 2 Preliminary Knowledge 2.1 Echo State Network 2.2 ESN-ADP 3 The Online Training Algorithm of ESN-ADP 3.1 The FORLS with Forgetting Parameter for ESN 3.2 The Online Training Process for ADP Controller 4 Experiment and Discussion 5 Conclusion References Deep Echo State Network Based Neuroadaptive Control for Uncertain Systems 1 Introduction 2 Deep Echo State Network 3 System Description and Problem Statement 4 Controller Design and Stability Analysis 5 Simulation 6 Conclusion References Bolt Loosening Detection Based on Principal Component Analysis and Support Vector Machine 1 Introduction 2 The Basic Principle of Wave-Guided Bolt Loosening Detection 3 Bolt Loosening Monitoring Method Based on PCA and SVM 3.1 PCA 3.2 SVM 3.3 Identification Method of Bolt Loosening Detection Based on SVM and PCA 4 Experimental Verification 4.1 Single-Bolt Connection Structure Ultrasonic Wave-Guiding Experiment 4.2 Dataset Creation 5 Experimental Results 5.1 Results of PCA 5.2 Comparison of Classification Results 5.3 Effect of the Number of Principal Components on the Classification Performance of PCA+SVM Method 5.4 Effect of Kernel Function on the Classification Performance of PCA+SVM Method 6 Conclusion References Detection and Identification of Digital Display Meter of Distribution Cabinet Based on YOLOv5 Algorithm 1 Introduction 2 Method 2.1 Instrument Character Area Detection and Positioning 2.2 Instrument Character Segmentation 2.3 Instrument Character Recognition 3 Experimental Results and Analysis 3.1 Experimental Platform 3.2 Acquisition and Labeling of Experimental Data Sets 3.3 Model Training 3.4 Results Analysis 4 Conclusion and Prospect References Analysis of Autoencoders with Vapnik-Chervonenkis Dimension 1 Introduction 2 Related Work 2.1 Statistics Concepts on VC-Dimension 2.2 Single-layer Network with VC-dimension 3 VC-dimension in Autoencoders 3.1 VC Dimension of Known Autoencoders 3.2 VC Dimension of Autoencoders with Unfixed Structure 4 Conclusion References Broad Learning with Uniform Local Binary Pattern for Fingerprint Liveness Detection 1 Introduction 2 Related Work 2.1 Hardware-Based FLD 2.2 Software-Based FLD 3 Methodology 3.1 Uniform Local Binary Pattern 3.2 Broad Learning System Model 4 Experiments and Results 5 Conclusion References A Novel Trajectory Tracking Controller for UAV with Uncertainty Based on RBF and Prescribed Performance Function 1 Introduction 2 System Model and Problem Statement 2.1 System Model 2.2 Problem Statement 3 Main Results 3.1 Prescribed Performance Control 3.2 Fractional Order Sliding Mode Control 3.3 The FO-SMC with PPC and RBF 4 Experiments Simulation 5 Conclusions References Item-Behavior Sequence Session-Based Recommendation 1 Introduction 2 Related Work 2.1 Session-Based and Sequential Recommendation 3 Methodology 3.1 Problem Formulation 3.2 Model Overview 3.3 Session Modeling 3.4 Session Representation Generating 4 Experiment 4.1 Experiment Settings 4.2 Global Performance Comparisons 4.3 Ablation Study 5 Conclusion References Human-Centered Real-Time Instance Segmentation with Integration with Data Association and SOLO 1 Background 2 Brief Review of Related Work 3 The Proposed Method 3.1 Overview 3.2 Feature Extraction Module 3.3 Similarity Estimation Module 3.4 Tracking Trajectory Generation 3.5 Training 4 Experiments 4.1 Simulation Settings and Dataset 4.2 Implementation Details 4.3 Comparison Results 5 Conclusion References Extracting Key Information from Shopping Receipts by Using Bayesian Deep Learning via Multi-modal Features 1 Introduction 2 Related Work 3 Key Information Extraction Based on Bayesian Deep Learning 3.1 Problem Formulation and Framework Overview 3.2 Feature Extraction 3.3 Bayesian Deep Network for Classification 3.4 Implementation Details 4 Experiment 4.1 Dataset 4.2 Experiment Settings and Evaluation Metrics 4.3 Results and Comparison 4.4 Ablation Study 5 Conclusion References A Multi-Surrogate-Assisted Artificial Bee Colony Algorithm for Computationally Expensive Problems 1 Introduction 2 Artificial Bee Colony Algorithm 3 Proposed Approach 3.1 Modified Search Strategy 3.2 Random Dimension Perturbation 3.3 Radial Basis Function Network 3.4 Kriging Model 3.5 Multi-Surrogate-Assisted ABC 3.6 Surrogate Model Management 4 Experimental Study 4.1 Test Problems and Parameter Settings 4.2 Computational Results 5 Conclusion References Multi-view Spectral Clustering with High-order Similarity Learning 1 Introduction 2 Related Work 3 Proposed Method 3.1 Main Notations 3.2 First and Second Order Similarity Learning 3.3 Multi-view Spectral Clustering with High-order Similarity Learning 3.4 Optimization 3.5 Complexity Analysis 4 Experiment 4.1 Datasets 4.2 Baselines and Evaluation Metrics 4.3 Results Analysis 4.4 Parameters Analysis 4.5 Convergence Analysis 5 Conclusion References An Improved Convolutional Neural Network Model by Multiwavelets for Rolling Bearing Fault Diagnosis 1 Introduction 2 Basic Theory of Method 2.1 Multiwavelets 2.2 CNN 3 The Presented Network Structure 3.1 Multiwavelets Layer 3.2 Multiwavelets Layer Parameters 4 Experiment Verification 4.1 Experiment Description 4.2 Parameter Optimization of Multiwavelets Layer 4.3 The Performance of Bearing Fault Diagnosis 5 Conclusion References TextSMatch: Safe Semi-supervised Text Classification with Domain Adaption 1 Introduction 2 Related Work 3 Method 3.1 Overview 3.2 OOD Data Detection 3.3 Adversarial Domain Adaption 4 Experiment 4.1 Dataset 4.2 Baseline Methods 4.3 The Settings 4.4 The Results 5 Conclusion References Recommendation Method of Cross-language Computer Courses 1 Introduction 2 Related Works 3 Recommendation Methods 3.1 Content-Based Recommendation Method 3.2 Cross-Language Recommendation Methods 3.3 Other Models 4 Experimental Setups and Results 4.1 Datasets 4.2 Fine-Tuning of the Course Classification Model 4.3 Experimental Evaluation 4.4 Recommendation Method Evaluation 4.5 Internal Evaluation of Translation Models 5 Conclusion References Temperature Prediction of Medium Frequency Furnace Based on Transformer Model 1 Introduction 2 Principle Introduction 2.1 Principle of Series Resonant Medium Frequency Furnace 2.2 Effect of Temperature on Metal Resistivity 2.3 Transformer Model 2.4 Simulation Verification of Circuit Model of Medium Frequency Furnace 3 Acquisition of Training Data of Transformer Model 3.1 Collection Principle of Equivalent Resistance of Charge Resistance in Primary Circuit 3.2 Construction of Equivalent Resistance Data Acquisition System 3.3 Collection of Charge Temperature Data 4 Using Transformer Model to Process Data 5 Construction of Burden Smelting Temperature Calculation System 6 Result Analysis 7 Conclusion References LQR Optimal Control Method Based on Two-Degree-of Freedom Manipulator 1 Introduction 2 Dynamical Modeling 3 LQR Controller 4 Genetic Algorithm 5 Simulation Results 6 Conclusion References Multi-layer Echo State Network with Nonlinear Vector Autoregression Reservoir for Time Series Prediction 1 Introduction 2 Echo State Network 2.1 Traditional Echo State Network 2.2 Multi-layer Echo State Network 3 MLESN with Nonlinear Vector Autoregression Reservoir 3.1 Nonlinear Vector Autoregression Reservoir 3.2 Multi-layer Nonlinear Vector Autoregression Reservoir 4 Time Series Prediction Simulation Outcomes 4.1 Experimental Data Presentation 4.2 Model Performance and Comparison 4.3 Discussion 5 Conclusion References Observer-Based Adaptive Security Control for Network Control Systems Under TDS Actuator Attacks 1 Introduction 2 Dynamic Model of NCS 3 State Feedback Control Design 3.1 Observer and Controller Design Under Normal Conditions 3.2 Observer and Controller Design Under TDS Attack 4 Stability for NCS with TDS Attack and Delay Detection Method 5 Output Feedback Control Design 5.1 Controller Design 5.2 Observer Design 6 Numerical Simulation 6.1 Vulnerability Analysis 6.2 TDS Attack Detection and State Estimaion 7 Conclusion References Feature Selection for High-Dimensional Data Based on a Multi-objective Particle Swarm Optimization with Self-adjusting Strategy Pool 1 Instruction 2 Related Works 2.1 PSO-Based Feature Selection 2.2 Strategy Pool 3 The Proposed Method 3.1 AMB Feature Prioritization Approach 3.2 Initialization 3.3 Self-adjusting Strategy Pool 3.4 Proposed Algorithm: MOPSO-SaSP 4 Experimental Results and Analysis 4.1 Datasets 4.2 Compared Algorithms and Parameter Setting 4.3 Performance of Feature Prioritization 4.4 Performance of Self-adjusting Strategy Pool 4.5 Performance of Multi-objective Frame 5 Conclusion and Future Work References Data-Driven Recommendation Model with Meta-learning Autoencoder for Algorithm Selection 1 Introduction 2 Background 2.1 Algorithm Selection 2.2 Meta-learning 2.3 Autoencoder 3 Data-Driven Recommendation Model 3.1 Meta-learning Module 3.2 Algorithm Recommending Module 4 Experiment and Results 4.1 Experimental Setup 4.2 Experimental Designs and Results 5 Conclusion References Author Index