Computational Science – ICCS 2022: 22nd International Conference, London, UK, June 21–23, 2022, Proceedings, Part II (Lecture Notes in Computer Science, 13351)

The four-volume set LNCS 13350, 13351, 13352, and 13353 constitutes the proceedings of the 22ndt International Conference on Computational Science, ICCS 2022, held in London, UK, in June 2022.* The total of 175 full papers and 78 short papers presented in this book set were carefully reviewed and selected from 474 submissions. 169 full and 36 short papers were accepted to the main track; 120 full and 42 short papers were accepted to the workshops/ thematic tracks. *The conference was held in a hybrid format Preface Organization Contents – Part II ICCS 2022 Main Track Short Papers Neuroevolutionary Feature Representations for Causal Inference 1 Introduction 2 Related Work 3 Methodology 4 Ablation Study on Generated Data 5 Application to Econometric Data 6 Conclusions References Compiling Linear Algebra Expressions into Efficient Code 1 Introduction 2 A Language for Linear Algebra Expressions 3 Transformation from Index Form into Index-Free Form 4 Compilation into Multiple Backends 5 Conclusion References Interval Modification of the Fast PIES in Solving 2D Potential BVPs with Uncertainly Defined Polygonal Boundary Shape 1 Introduction 2 Modelling Uncertainty of the Boundary Shape 3 The Interval Modified Fast PIES (IFPIES) 4 Numerical Results 5 Conclusions References Validation and Optimisation of Player Motion Models in Football 1 Introduction 2 Motion Models in Football: State of the Art 3 Player Motion Model and Validation Procedure 3.1 Measuring Correctness 3.2 Measuring Precision 3.3 Defining an Overall Validation Score 4 Experiment and Evaluation of Results 4.1 Data Set 4.2 Preparation of Motion Models 4.3 Evaluation of Results 5 Conclusion References Auto-scaling of Scientific Workflows in Kubernetes 1 Introduction 2 Related Work 3 Auto-scaling Scientific Workflows in Kubernetes 3.1 Monitoring 3.2 Analysis 3.3 Planning 3.4 Execution 3.5 Autoscaling Policies – Reactive vs. Predictive 4 Evaluation 4.1 Experiment Setup 4.2 Results 5 Conclusions and Future Work References Transfer Learning Based Natural Scene Classification for Scene Understanding by Intelligent Machines 1 Introduction 2 Literature Review 3 Proposed Work 4 Transfer Learning-ResNet 5 Proposed Methodology 5.1 Dataset 5.2 Pre-trained Models 6 Experimental Results 6.1 Training Progress - ResNet50 and ResNet 101 6.2 Performance Metrics 7 Conclusion References Devulgarization of Polish Texts Using Pre-trained Language Models 1 Introduction 2 Theoretical Background for the Corpus Work 3 Training Data 4 Devulgarization Tool 4.1 GPT-2-Based Solution 4.2 GPT-3-Based Solution 4.3 T5-Based Solution 5 Evaluation 5.1 Evaluation Method 5.2 Baselines 5.3 Quantitative Evaluation Results 5.4 Qualitative Evaluation Results 6 Conclusions and Future Work References Data-Driven Discovery of Time Fractional Differential Equations 1 Introduction 2 Methodology 2.1 Sparse Reconstruction of Fractional Differential Equation 2.2 Differential Evolution 3 Simulations and Numerical Experiments 4 Conclusions References Wide Ensembles of Neural Networks in Music Genre Classification 1 Introduction 2 Related Work 3 Conditions of Experiments 4 Experiments 4.1 Basic Dense Classifiers 4.2 Principal Component Analysis Influence 4.3 Influence of the Number of Classifiers 4.4 Influence of the Architecture of Base-Level Classifiers 5 Comparison of the Outcomes 6 Conclusions References MultiEmo: Language-Agnostic Sentiment Analysis 1 Introduction 2 Related Work 3 Language-agnostic Models Combined with LSTM 4 Experimental Setup 4.1 Pipeline 4.2 MultiEmo Dataset 4.3 Scenarios 5 Experimental Results 6 Conclusion and Future Work References Particle Swarm Optimization Configures the Route Minimization Algorithm 1 Introduction 1.1 Problem Formulation 1.2 Related Work 2 Configuring the Guided Ejection Search Using PSO 3 Experiments 4 Conclusions References Practical Aspects of Zero-Shot Learning 1 Introduction and Literature Review 2 Selection of Methods and Experimental Setup 3 Experiments with Individual Classifiers 3.1 Analysis of the Datasets 4 Meta-Classifiers 5 Concluding Remarks References A Hypothetical Agent-Based Model Inspired by the Abstraction of Solitary Behavior in Tigers and Its Employment as a Chain Code for Compression 1 Introduction 2 Proposed Algorithm 3 Experimental Results and Analysis 4 Conclusion and Future Work References Analyzing the Usefulness of Public Web Camera Video Sequences for Calibrating and Validating Pedestrian Dynamics Models 1 Introduction 2 Application of Web Cameras 2.1 Object Identification 2.2 Mapping from 2D Camera Space to 3D World Space 3 Conclusions References A Highly Customizable Information Visualization Framework 1 Introduction 2 Background 3 Solution Conceptualization 4 Implementation Details 5 Experimental Setup and Results 6 Conclusions and Future Work References Incremental Dynamic Analysis and Fragility Assessment of Buildings with Different Structural Arrangements Experiencing Earthquake-Induced Structural Pounding 1 Introduction 2 Incremental Dynamic Analysis and Fragility Assessment 3 Numerical Models of Buildings 4 IDA 5 Fragility Assessment 6 Conclusion References PIES with Trimmed Surfaces for Solving Elastoplastic Boundary Problems 1 Introduction 2 PIES for Elastoplastic Problems 3 The Plastic Zone Modeled by the Trimmed Surface 4 Numerical Solving of PIES with the Trimmed Surface 5 Examples 6 Conclusions References Linear Computational Cost Implicit Variational Splitting Solver with Non-regular Material Data for Parabolic Problems 1 Introduction 2 Varying Coefficients in Alternating Directions Solver 3 Variational Splitting for Pennes Bioheat Equation References A Hadamard Matrix-Based Algorithm to Evaluate the Strength of Binary Sequences 1 Introduction 2 Preliminaries 3 Binomial Representation of Binary Sequences 4 An Algorithm to Compute the LC of Binary Sequences 4.1 Application of the Algorithm to Generalized Sequences 4.2 Comparison with Other Algorithms 5 Conclusions References Private and Public Opinions in a Model Based on the Total Dissonance Function: A Simulation Study 1 Introduction 2 The Model 3 Monte Carlo Simulations 4 Results 5 Conclusions References Analysis of Public Transport (in)accessibility and Land-Use Pattern in Different Areas in Singapore 1 Introduction 2 Data and Methods 3 Results and Discussion 4 Conclusion References Pseudo-Newton Method with Fractional Order Derivatives 1 Introduction 2 Fractional Derivatives 3 The Pseudo-Newton Method 4 The Pseudo-Newton Method with Fractional Derivatives 5 Numerical Results 6 Conclusions References An Energy Aware Clustering Scheme for 5G-Enabled Edge Computing Based IoMT Framework 1 Introduction 2 System Model and Assumptions 3 Experimental Results and Discussion 4 Concluding Remarks References A Framework for Network Self-evolving Based on Distributed Swarm Intelligence 1 Introduction 2 Model of Self-evolving Network 3 Self-evolving Architecture 3.1 Distributed Swarm Intelligence Algorithm 3.2 Consensus Search 3.3 Topology Self-evolving 4 Performance Evaluation 5 Conclusion References Investigating an Optimal Computational Strategy to Retrofit Buildings with Implementing Viscous Dampers 1 Introduction 2 Modeling Approach 3 Computational Retrofitting Strategy 4 Results of Strategy 5 Conclusion References ARIMA Feature-Based Approach to Time Series Classification 1 Introduction 2 Literature Review 3 The Method 4 Results 5 Conclusion References A Note on Adjoint Linear Algebra 1 Motivation 2 Prerequisites 3 BLAS Revisited 4 Systems of Linear Equations Revisited 5 Conclusion References Approximate Function Classification 1 Introduction 2 Experiments 2.1 Datasets 2.2 Four-Variable Irreversible Logic Functions 2.3 From n-1 to n Classification 3 NPN Classes Discrimination 4 Conclusion References Acceleration of Optimized Coarse-Grid Operators by Spatial Redistribution for Multigrid Reduction in Time 1 Introduction 2 Multigrid Reduction in Time 3 Spatial Redistribution Technique 4 Numerical Experiments 5 Conclusion References Networks Clustering-Based Approach for Search of Reservoirs-Analogues 1 Introduction 2 Algorithms and Methods 2.1 Networks Clustering Algorithms 2.2 Search for a Cluster of Reservoirs-Analogues for the Target Reservoir 3 Experiments and Results 3.1 Using Different Distance Metrics 3.2 Using Different Clusters Detection Algorithms 3.3 Methods for Parameters Recovery 4 Conclusion References KP01 Solved by an n-Dimensional Sampling and Clustering Heuristic 1 Introduction 2 Description of the MCDKM Heuristic Method 3 The Knapsack Problem into the MCDKM Method 4 Implementation and Empirical Results 5 Conclusions and Future Work References A Deep Neural Network as a TABU Support in Solving LABS Problem 1 Introduction 2 Tabu Search for LABS Problem 3 Improving Tabu Search with LSTM 4 Effectiveness of the Proposed Algorithms 5 Conclusions and Further Work References Classification and Generation of Derivational Morpho-semantic Relations for Polish Language 1 Introduction 2 Related Work 3 Data 4 Proposed Approach and Experiments 5 Discussion 6 Conclusion and Future Work References On the Explanation of AI-Based Student Success Prediction 1 Introduction and Background 2 Methodology 2.1 Student Success Prediction 2.2 Explaining the Predictions 3 Experimental Setup 3.1 The Dataset 3.2 Prediction Results 3.3 Prediction Explanation 4 Discussion and Analysis 4.1 Investigating User Perceptions 5 Conclusions and Future Work References The Importance of Scaling for an Agent Based Model: An Illustrative Case Study with COVID-19 in Zimbabwe 1 Introduction 2 Methodology 2.1 Overview 2.2 Design 2.3 Details 3 Data 3.1 Synthetic Populations 4 Results 5 Discussion and Conclusion References Phase-Field Modelling of Brittle Fracture Using Time-Series Forecasting 1 Introduction 2 Isotropic Brittle Fracture Model 2.1 Phase-Field Approximation 2.2 Treating the Crack Propagation Process as a Time-Series 3 Time-Series Forecasting 3.1 Statistics Based Time-Series Forecasting with ARIMA 3.2 Deep-Learning Based Time-Series Modelling with LSTM 4 Case Study - Single Edge Notched Tensile 5 Results 5.1 Prediction Accuracy Rate 5.2 Runtime Performance of the Prediction Models 6 Conclusion and Future Work References Advances in High-Performance Computational Earth Sciences: Applications and Frameworks Calculation of Cross-correlation Function Accelerated by Tensor Cores with TensorFloat-32 Precision on Ampere GPU 1 Introduction 2 Methodology 2.1 Usage of Tensor Cores with TF32 2.2 Calculation of Cross-correlation Function Using Tensor Cores 3 Application and Performance Measurement 3.1 Application Example 3.2 Performance Measurement 4 Closing Remarks References Developing an ELM Ecosystem Dynamics Model on GPU with OpenACC 1 Introduction 1.1 Computational Platform and Software Environment 2 Method 2.1 ELM Data Structure and Ecosystem Dynamics Model 2.2 Performance Comparison and a Naive OpenACC Implementation 2.3 Code Optimization 3 Results 3.1 Overall Performance Improvement 3.2 Profiling Details 4 Conclusion and Future Work References Artificial Intelligence and High-Performance Computing for Advanced Simulations Optimization-Free Inverse Design of High-Dimensional Nanoparticle Electrocatalysts Using Multi-target Machine Learning 1 Introduction 2 Methods 2.1 Data Set 2.2 Multi-target Random Forest Regression 2.3 Workflow 3 Discussion of Results 4 Conclusions References CNNs with Compact Activation Function 1 Introduction 2 Hat Function for MgNet 3 Experiments 4 Conclusion References Deep Neural Networks and Smooth Approximation of PDEs 1 Introduction References Isogeometric Analysis of Bound States of a Quantum Three-Body Problem in 1D 1 Introduction 2 The Two- And Three-Body Problems 2.1 The Unified Problem 2.2 A Solution Example of the Two-Body Problem 3 Isogeometric Analysis 3.1 Continuous Level 3.2 IGA Discretized Level 3.3 Algebraic Level 3.4 A Priori Error Estimates 4 Numerical Experiments 4.1 IGA Discretization Accuracy 4.2 A Study on Domain Size 4.3 Three-Body Problem 5 Concluding Remarks References 1D Painless Multi-level Automatic Goal-Oriented h and p Adaptive Strategies Using a Pseudo-Dual Operator 1 Introduction 2 Proposed Goal-Oriented Adaptive Algorithms 2.1 Error Indicators 3 Numerical Results 3.1 Helmholtz Goal-Oriented Problem 3.2 Convection-Dominated Goal-Oriented Problem 4 Conclusions References Transfer Learning Approach to Prediction of Rate of Penetration in Drilling 1 Introduction 2 Transfer Learning 3 Experimental Setup and Data 3.1 Methodology 3.2 Datasets 3.3 Setup of Experiments 3.4 Model Evaluation 3.5 Example of Usage 4 Results 4.1 3-way-comparison 4.2 Recommendations Based on All Test Cases 4.3 Test Outside Volve Data 5 Conclusions References Physics Informed RNN-DCT Networks for Time-Dependent Partial Differential Equations 1 Introduction 2 Methods 3 Experiments 4 Conclusion References Recursive Singular Value Decomposition Compression of Refined Isogeometric Analysis Matrices as a Tool to Speedup Iterative Solvers Performance 1 Introduction 2 Matrices of Refined Isogeometric Analysis 3 Compression Algorithm 4 Multiplication Algorithm 5 Results 6 Summary of the Results 7 Conclusions and Future Work References Neural-Network Based Adaptation of Variation Operators' Parameters for Metaheuristics 1 Introduction 2 Neural Networks in Prediction and Control 3 Adaptation of Evolutionary Search Parameters 4 ANN-Based Adaptation of Evolutionary Search 5 Experimental Results 6 Conclusion References Performance of Computing Hash-Codes with Chaotically-Trained Artificial Neural Networks 1 Introduction 2 Related Work 3 The Chaotic Model Used 4 ANNs Training and Testing with Usage of Lorenz Attractor 5 Analysis of Performance of the Hashing ANNs 6 Conclusions References Application of the Hierarchic Memetic Strategy HMS in Neuroevolution 1 Introduction 2 Hierarchic Memetic Strategy (HMS) 3 Application in Neuroevolution 4 Experiments 5 Conclusions References Biomedical and Bioinformatics Challenges for Computer Science CXR-FL: Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning 1 Introduction 2 Method: FL for Segmentation and Classification 3 Experiments and Results 3.1 Datasets 3.2 Implementation Details 3.3 Segmentation Results 3.4 Classification Results 4 Conclusions References POTHER: Patch-Voted Deep Learning-Based Chest X-ray Bias Analysis for COVID-19 Detection 1 Introduction 2 Method 3 Experiments and Results 3.1 Datasets 3.2 Data Pre-processing 3.3 Implementation Details 3.4 Gradient-Weighted Class Activation Mapping Results 3.5 Quantitative Analysis 4 Discussion 5 Conclusions References Modeling Contrast Perfusion and Adsorption Phenomena in the Human Left Ventricle 1 Introduction 2 Methods 2.1 Mathematical Modeling 2.2 Left Ventricular Geometry Models 2.3 Numerical Methods 3 Results and Discussion 3.1 Simplified LV Model Study 3.2 Patient Specific LV Study 4 Conclusions References Dense Temporal Subgraphs in Protein-Protein Interaction Networks 1 Introduction 2 Definitions 3 An Exact Algorithm for k-Densest-Temporal-Refinement 4 A Heuristic of k-Densest-Temporal-Refinement 5 Experimental Analysis 6 Conclusion References Accelerating Edge Metagenomic Analysis with Serverless-Based Cloud Offloading 1 Introduction 2 Related Works 3 Testing Workflow and Environment 4 Performance Experiments 5 Results Summary and Concluding Remarks References Continuous-to-Continuous Data Model vs. Discrete-to-Discrete Data Model for the Statistical Iterative Reconstruction Method 1 Introduction 1.1 Motivation 1.2 Contribution 2 Statistical Reconstruction Approaches 2.1 A Statistical Approach Based on the Discrete-to-Discrete Data Model 2.2 A Statistical Approach Based on the Continuous-to-Continuous Data Model 3 Experimental Results 4 Conclusion References Musculoskeletal Model of Human Lower Limbs in Gait Simulation 1 Introduction 2 Methodology 2.1 Skeletal Model 2.2 Muscle Modeling 2.3 3D Fiber Representation 2.4 Muscle Model 2.5 Combination of Muscle Module with Skeletal Biomechanical Module 2.6 Results 3 Conclusions References Tissue Damage Control Algorithm for Hyperthermia Based Cancer Treatments 1 Introduction 2 Methods 2.1 Mathematical Model 2.2 Numerical Scheme 2.3 Differential Evolution 3 Results 4 Discussion 5 Conclusions and Future Works References Resting-State EEG Classification for PNES Diagnosis 1 Introduction 2 Methods 2.1 EEG Pre-processing 2.2 EEG Features Extraction 2.3 EEG Classification 3 Results 4 Conclusions References Machine Learning Approaches in Inflammatory Bowel Disease 1 Introduction 2 Data Sources in IBD 2.1 IBD Databases 3 Need for Machine Learning in IBD 4 Discussion 5 Conclusions and Future Work References A Machine Learning Framework for Fetal Arrhythmia Detection via Single ECG Electrode 1 Introduction 2 Materials and Methods 2.1 Discrete Wavelet Transform (DWT) 2.2 Data Description 2.3 Proposed Model 2.4 Performance Metrics 3 Results 4 Conclusion References Computational Collective Intelligence Consensus Algorithm for Bi-clustering Analysis 1 Introduction 2 Description of the Proposed Consensus Algorithm 2.1 Notation 2.2 Assignment Step 2.3 Trimming Step 3 Comparisons of Performances of Bi-clustering Algorithms 3.1 Synthetic Data 3.2 Text Mining Data 4 Results and Discussion 4.1 Synthetic Data 4.2 Real Data 5 Conclusions References Collective of Base Classifiers for Mining Imbalanced Data 1 Introduction 2 Related Work 3 GEP-NB Classifier 3.1 General Idea of the GEP-NB 3.2 Formal Description of the Approach 4 Computational Experiment 5 Comparative Analysis 6 Conclusion References Impact of Clustering on a Synthetic Instance Generation in Imbalanced Data Streams Classification 1 Introduction 2 A Framework for Learning from Imbalanced Data Streams 3 Over-Sampling Procedure for Imbalanced Data Streams 3.1 Basic Over-Sampling Procedure 3.2 A Modified Procedure for Over-Sampling 4 Computational Experiment 5 Conclusions References Enhancing Decision Combination in Classifier Committee via Positional Voting 1 Introduction 2 Algorithm 2.1 Main Ingredients of Classifier Committee 2.2 Dynamic Time Warping-Based Action Features 2.3 Optimized Positional Voting 3 Experimental Results 4 Conclusions References Competition and Cooperation Mechanisms for Collective Behavior in Large Multi-agent Systems 1 Introduction 2 Iterated Spatial Prisoner's Dilemma Game and Collective Behavior 3 CA–Based Players 4 Competition and Income Sharing Mechanisms 5 Experimental Results 5.1 Competition in Systems with a Single Adaptive Strategy 5.2 Competition in Heterogeneous Systems 5.3 Competition and Cooperation: A Case Study 6 Conclusions References Divergence of an Observed User Profile and a Simulated Real State of User Due to Social Communication 1 Introduction 2 Related Works 3 Group Models 3.1 Model of the Social Collective 3.2 Model of the Collaborative Filtering Group 3.3 Combining the Models 4 Evaluation of the Model 5 Conclusions References Temporal-Attribute Inference Using Dynamic Bayesian Networks 1 Introduction 2 Related Work 3 Methodology 4 Features 4.1 Second-Layer Features 4.2 Network Features 5 Temporal Modeling Using Dynamic Bayesian Networks 5.1 Feature Selection and Model Selection 5.2 Intention-Specific Models 6 Data Collection 7 Experimental Results References Fuzzy Logic Framework for Ontology Instance Alignment 1 Introduction 2 Related Works 3 Basic Notions 4 Fuzzy Based Approach to Instance Alignment 5 Experimental Verification 6 Future Works and Summary References Neuro-Symbolic Models for Sentiment Analysis 1 Introduction 2 Related Work 3 Datasets 3.1 plWordNet and plWordNet Emo 3.2 PolEmo 3.3 Preprocessing 4 Neuro-Symbolic Models 4.1 HB: HurtBERT Model 4.2 TK: Tailored KEPLER Model 4.3 TE: Token Extension Model 4.4 ST(P): Special Tokens (with Positioning) Model 4.5 STN: Special Tokens from Numeric Data Model 4.6 SE: Sentiment Embeddings Model 5 Experimental Setup and Results 5.1 TK: Tailored KEPLER Model 5.2 TE: Token Extension Model 5.3 ST(P): Special Tokens (with Positioning) Model 5.4 STN: Special Tokens from Numeric Data 5.5 HB: HurtBERT Model and SE: Sentiment Embeddings Model 6 Conclusions References A Unified Sense Inventory for Word Sense Disambiguation in Polish 1 Introduction 2 Related Work 3 Polish WordNet as a Knowledge Base 4 Polish WSD Inventory 5 Evaluation 6 Conclusions References Sentence-level Sentiment Analysis Using GCN on Contextualized Word Representations 1 Introduction 2 Related Works 3 Research Problem 3.1 Problem Definition 3.2 Research Questions 4 Proposed Method 4.1 Creating Sentence Vectors Based on BERT 4.2 Creating the Contextualized Word Representations 4.3 Building the Contextual-Based GCN Model 4.4 Sentence-level Sentiment Classifier 5 Experimental Evaluation 5.1 Dataset and Experimental Setup 5.2 Baseline Methods 5.3 Results and Discussion 6 Conclusion and Future Works References Machine Learning for Bus Travel Prediction 1 Introduction and Related Work 2 Data, Its Preprocessing, and Experimental Setup 3 Experimental Results and Their Analysis 4 Concluding Remarks References Prediction of Ether Prices Using DeepAR and Probabilistic Forecasting 1 Introduction 2 Background and Related Works 2.1 Background 2.2 Related Works 3 Experiments and Results 3.1 Data Collection and Pre-processing 3.2 Experimental Setup 3.3 Experimental Results 3.4 Discussion 4 Conclusions and Future Works References Purchasing Decisions on Alternative Fuel Vehicles Within an Agent-Based Model 1 Introduction 2 The Model 3 Results 4 Discussion References Author Index