Artificial Intelligence Applications and Innovations

Name: Artificial Intelligence Applications and Innovations | 15th IFIP WG 12.5 International Conference, AIAI 2019, Hersonissos, Crete, Greece, May 24-26, 2019, Proceedings
Brand: Springer
Price: 96.29 EUR
Availability: OnlineOnly

15th IFIP WG 12.5 International Conference, AIAI 2019, Hersonissos, Crete, Greece, May 24-26, 2019, Proceedings

John MacIntyre Ilias Maglogiannis Lazaros Iliadis Elias Pimenidis(Herausgeber*in)

Springer (Verlag)

Erschienen am 15. Mai 2019

XXIV, 689 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-030-19823-7 (ISBN)

96,29 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Intro
Preface
AIAI 2019
Organization
Abstracts of Invited Talks
Learning from Electronic Health Records: From Temporal Abstractions to Time Series Interpretability
Empirical Approach: How to Get Fast, Interpretable Deep Learning
"In-memory Computing": Accelerating AI Applications
Contents
Invited Paper
The Power of the ``Pursuit'' Learning Paradigm in the Partitioning of Data
1 Introduction
2 The Object Migration Automata
3 Developing the Pursuit Concept: The Environment
3.1 The Design and Results of the POMA
4 Enhanced OMA (EOMA)
5 Enhancing the EOMA with a Pursuit Paradigm
6 Cohesiveness in the EPP: The Transitive PEOMA
7 Conclusions
References
AI Anomaly Detection - Active Learning
Cyber-Typhon: An Online Multi-task Anomaly Detection Framework
Abstract
1 Introduction
2 Literature Review
3 The Proposed Cyber-Typhon Framework
4 Datasets
4.1 Training the RBMs
5 Results
6 Discussion and Conclusions
References
Investigating the Benefits of Exploiting Incremental Learners Under Active Learning Scheme
Abstract
1 Introduction
2 Related Work
2.1 Incremental Learning
2.2 Active Learning
3 Proposed Framework
4 Experimental Procedure and Results
5 Conclusions and Future Work
Acknowledgements
References
The Blockchain Random Neural Network in Cybersecurity and the Internet of Things
Abstract
1 Introduction
1.1 Research Motivation
1.2 Research Proposal
2 Related Work
2.1 Internet of Things
2.2 Neural Networks in Cryptography
2.3 Blockchain in Security
3 Blockchain Neural Network in the Internet of Things
3.1 The Random Neural Network
3.2 The Random Neural Network with Blockchain Configuration
4 Cybersecurity and the IoT Blockchain Model
4.1 Private Key
4.2 Roaming and Verification
4.3 Neural Chain Network and Mining
4.4 Decentralized Information
5 Neural Blockchain in Cybersecurity and IoT Validation
6 Conclusions
Appendix
References
Autonomous Vehicles - Aerial Vehicles
A Visual Neural Network for Robust Collision Perception in Vehicle Driving Scenarios
1 Introduction
2 Model Description
2.1 Spatiotemporal Neural Computation in the Pre-synaptic Area
2.2 Adaptive Inhibition Mechanism
2.3 The LGMD Cell
2.4 Output Spike Frequency
3 Experimental Evaluation
3.1 Synthetic Visual Stimuli Testing
3.2 Real-World Driving Scenes Testing
3.3 Discussion
4 Concluding Remarks
References
An LGMD Based Competitive Collision Avoidance Strategy for UAV
1 Introduction
2 Model Description
2.1 LGMD Process
3 System Overview
3.1 Quadcopter Platform
4 Experiments and Results
4.1 Video Simulation
4.2 Hovering and Features Analysis
4.3 Arena Real-Time Flight
5 Conclusion
References
Mixture Modules Based Intelligent Control System for Autonomous Driving
1 Introduction
2 Related Work
3 Lateral and Longitudinal Controller
3.1 Lateral Controller
3.2 Longitudinal Controller
4 Experiments and Results
4.1 Experimental Platform
4.2 Path Tracking Experiment
4.3 Speed Control Experiment
5 Conclusion
References
Biomedical AI
An Adaptive Temporal-Causal Network Model for Stress Extinction Using Fluoxetine
Abstract
1 Introduction
2 Underlying Neurological Principles
3 The Temporal-Causal Network Model
4 Example Simulation
5 Mathematical Analysis
6 Conclusion
References
Clustering Diagnostic Profiles of Patients
1 Introduction
2 Patient Data and Diagnoses as Profiles
3 Experiments: Methodology and Evaluation
4 Summary and Conclusions
References
Emotion Analysis in Hospital Bedside Infotainment Platforms Using Speeded up Robust Features
Abstract
1 Introduction
2 Related Work
3 System Architecture
3.1 Overview
3.2 The Emotion Analysis Process
4 The System in Practice
5 Experimental Results
6 Conclusion
Acknowledgment
References
FISUL: A Framework for Detecting Adverse Drug Events from Heterogeneous Medical Sources Using Feature Importance
1 Introduction
2 Related Work
3 The FISUL Framework
3.1 Phase I: Boruta Feature Importance
3.2 Phase II: Predictive Modeling
3.3 Phase III: Clustering
4 Experimental Evaluation
5 Conclusions
References
Classification - Clustering
A New Topology-Preserving Distance Metric with Applications to Multi-dimensional Data Clustering
Abstract
1 Introduction
2 Methodology
2.1 The Proposed Topology-Preserving Distance Metric
2.2 Determining the Value of d0 Parameter
2.3 A k-Means Variant for the Proposed Topology-Preserving Distance Metric
3 Results
4 Conclusions
Acknowledgement
References
Classification of Incomplete Data Using Autoencoder and Evidential Reasoning
1 Introduction
2 Basics of Evidential Reasoning
3 Proposed algorithm
3.1 The Architecture of the Autoencoder
3.2 The Learning
3.3 Decision Making for a Test Point
4 Experiments
4.1 Experimental Set Up
4.2 Experiment 1
4.3 Experiment 2
4.4 Experiment 3
5 Conclusion
References
Dynamic Reliable Voting in Ensemble Learning
1 Introduction
2 Related Work
2.1 Supervised Learning as Base Classifiers
2.2 Confidence Score in Ensemble Learning
3 Dynamic Reliable Voting Algorithm
3.1 Problem Formulation and Modeling
3.2 Classifier Selection
4 Experimentation Results and Discussion
4.1 Data Description and Protocol
4.2 Results and Discussion
5 Conclusion
References
Extracting Action Sensitive Features to Facilitate Weakly-Supervised Action Localization
1 Introduction
2 Related Works
3 Action Sensitive Network
3.1 Action Sensitive Extractor
3.2 Action Detection Criterion
3.3 Network Details
4 Experiments
4.1 Dataset
4.2 Implementation Details
4.3 Ablation Study
4.4 Experiments on AutoLoc
5 Conclusion
References
Image Recognition Based on Combined Filters with Pseudoinverse Learning Algorithm
Abstract
1 Introduction
2 Related Work
3 Proposed Methodology
3.1 Gabor Kernel
3.2 Random Kernel
3.3 Pseudoinverse Kernel
3.4 Ensemble Model
4 Performance Evaluation
4.1 MNIST Dataset
4.2 CIFAR-10 Dataset
4.3 Discussion
5 Conclusions
Acknowledgements
References
Constraint Programming - Brain Inspired Modeling
Design-Parameters Optimization of a Deep-Groove Ball Bearing for Different Boundary Dimensions, Employing Amended Differential Evolution Algorithm
Abstract
1 Introduction
2 Mathematical Modelling of Ball Bearing Design Problem
2.1 Design Parameters
2.2 Objective Function
2.3 Constraints
3 Amended Differential Evolution Algorithm (ADEA)
4 Results and Discussion
5 Conclusions
References
Exploring Brain Effective Connectivity in Visual Perception Using a Hierarchical Correlation Network
1 Introduction
2 Monolayer Correlation Network
3 Hierarchical Correlation Network
3.1 Voxel Correlation in the HcorrNet
3.2 Corresponding Correlation in the HcorrNet
3.3 Stimulus-Sensitivity Activation Patterns
3.4 Forward Encoding Process Representation in the HcorrNet
4 Decoding Process
5 Experiments
6 Conclusion
References
Solving the Talent Scheduling Problem by Parallel Constraint Programming
1 Introduction
2 Preliminaries
3 A CSP Model
4 Solving the TS in Parallel
5 Numerical Results
6 Conclusion
References
Deep Learning - Convolutional ANN
A Deep Reinforcement Learning Approach for Automated Cryptocurrency Trading
1 Introduction
1.1 Related Work
2 Cryptocurrency and Bitcoin
2.1 Automated Trading
3 Reinforcement Learning
4 Q-learning Trading System
5 Experimental Data and Results
5.1 Bitcoin Historical Data
5.2 Results
6 Conclusions and Future Work
References
Capacity Requirements Planning for Production Companies Using Deep Reinforcement Learning
Abstract
1 Introduction
2 Methodology for Applying Deep Reinforcement Learning to Planning Processes
2.1 Step 1: Model Business Process
2.2 Step 2: Identify Relevant Planning Process Steps
2.3 Step 3: Convert Planning Process into Markov Decision Process
2.4 Step 4: Develop Reinforcement Learning Environment
2.5 Step 5: Configure Reinforcement Learning Agent
2.6 Step 6: Evaluate Learning Results
3 Use Case for Deep Planning Methodology (DPM)
3.1 Standard Business Process Model for Production Companies
3.2 Capacity Requirements Planning Process
3.3 Markov Decision Process for Capacity Requirements Planning
3.4 Learning Environment for Capacity Requirements Planning
3.5 Learning Agent for Shift Schedule
3.6 Experimental Evaluation
4 Conclusion and Future Work
References
Comparison of Neural Network Optimizers for Relative Ranking Retention Between Neural Architectures
Abstract
1 Introduction
2 Related Work
3 Methodology
3.1 Evolving Architectures
3.2 Experimental Setup
3.3 Implementation
4 Results
4.1 Neural Architectures
4.2 Monte-Carlo Simulations
5 Limitations
6 Conclusions
7 Future Work
Acknowledgements
References
Detecting Violent Robberies in CCTV Videos Using Deep Learning
1 Introduction
2 Proposed Method
2.1 UNI-Crime Dataset
2.2 Neural Network Training
3 Results
4 Conclusion
References
Diversity Regularized Adversarial Deep Learning
1 Introduction
2 Related Work
3 Method
3.1 Feature Diversification in GAN
4 Experiments
5 Conclusion
References
Interpretability of a Deep Learning Model for Rodents Brain Semantic Segmentation
1 Introduction
2 Machine Learning Interpretability
3 Methods
3.1 Machine Learning System
3.2 Interpretability System
4 Experiments
4.1 Database
4.2 Rodents' Brain Semantic Tissue Segmentation
4.3 Background Removal
4.4 Results
5 Conclusion
References
Learning and Detecting Stuttering Disorders
1 Introduction
2 Preliminaries
3 The Proposed Architecture
4 Detection Technique
4.1 Noise Removal
4.2 Audio File Preparation
4.3 Feature Extraction
4.4 Fragment Classification
4.5 Segment Classification
5 Experiments
6 Conclusions
References
Localization of Epileptic Foci by Using Convolutional Neural Network Based on iEEG
1 Introduction
2 Dataset
3 Method
3.1 Short Time Fourier Transform (STFT)
3.2 Z-Score Normalization
3.3 Convolutional Neural Network (CNN)
4 Experimental Result and Discussion
5 Conclusion
References
Review Spam Detection Using Word Embeddings and Deep Neural Networks
Abstract
1 Introduction
2 Review Spam Detection - A Literature Review
3 Dataset
4 Methods
5 Experimental Results
6 Conclusion
Acknowledgments
References
Tools for Semi-automatic Preparation of Training Data for OCR
1 Introduction
2 Related Work
3 Tools for Semi-automatic Annotation
3.1 Line Image Extraction
3.2 Character Segmentation
3.3 CRNN Classifier
3.4 Classifier Training
3.5 Line Annotator
4 Conclusion and Future Work
References
Training Strategies for OCR Systems for Historical Documents
1 Introduction
2 CRNN Classifier
3 Dataset
4 Synthetic Data Creation
4.1 Synthetic Data Generator
5 Training Strategies
6 Experiments
6.1 Binarization and White-Space-Padding Experiment
6.2 Annotated Dataset Experiment - Strategy 1
6.3 Synthetic Dataset Experiment - Strategy 2
6.4 Synthetic and Annotated Dataset Experiment - Strategy 3
7 Conclusions
References
A Review on the Application of Deep Learning in Legal Domain
Abstract
1 Introduction
2 Research Methodology
2.1 Literature Selection
2.2 Research Questions
3 Literature Review
3.1 Legal Data Search
3.2 Legal Data Analytics
3.3 Legal Intelligent Interfaces
4 Conclusion
References
Long-Short Term Memory for an Effective Short-Term Weather Forecasting Model Using Surface Weather Data
Abstract
1 Introduction
2 Weather Research and Forecasting Model
3 Proposed Deep Model Using Long Short-Term Memory (LSTM) Network
4 Methodology
4.1 Surface Weather Parameters
4.2 Data Collection and Preparation
4.3 Proposed Model with Optimal Number of LSTM Layers
5 Results and Discussion
6 Conclusion and Future Work
Acknowledgement
References
Segmentation Methods for Image Classification Using a Convolutional Neural Network on AR-Sandbox
Abstract
1 Introduction
2 Background
3 General Approach
4 CNN Base Model
5 Image Processing
5.1 Canny Edge Detector
5.2 Color-Space
5.3 Threshold
5.4 Similarity Indexes
6 Results
6.1 ROC Curves
6.2 Confusion Matrix
7 Analysis of Results
8 Conclusions
References
Fuzzy Modeling
A Hybrid Model Based on Fuzzy Rules to Act on the Diagnosed of Autism in Adults
1 Introduction
2 Literature Review
2.1 Autistic Spectrum Disorder (ASD)
2.2 Fuzzy Neural Networks: Principal Concepts
3 Pruning Fuzzy Neural Network for Detection of Autistic Persons
3.1 Network Architecture
3.2 Pruning Method - F-Score
3.3 Training of the Model Based on Extreme Learning Machine
4 Tests
4.1 Assumptions and Initial Test Configurations
4.2 Dataset Used in the Tests.
4.3 Binary Pattern Classification Tests
5 Conclusion
References
An Unsupervised Fuzzy Rule-Based Method for Structure Preserving Dimensionality Reduction with Prediction Ability
1 Introduction
2 Proposed Method
2.1 The Rule-Based System
2.2 Initial Rule Base Formation
2.3 Training the Rule Base
2.4 Rejection of Outputs
3 Experimentation
3.1 Dataset Descriptions and Experimental Settings
3.2 Results and Discussions
4 Conclusion
References
Interpretable Fuzzy Rule-Based Systems for Detecting Financial Statement Fraud
1 Introduction
2 Financial Statement Fraud Detection - A Literature Review
3 Research Methodology
3.1 Dataset
3.2 Feature Selection
3.3 Fuzzy Rule-Based Systems
3.4 Performance Evaluation
4 Experimental Results
5 Conclusion
References
Learning Automata - Logic Based Reasoning
Learning Automata-Based Solutions to the Single Elevator Problem
1 Introduction
1.1 Prior Art for the SEP
1.2 Learning Automata (LA)
1.3 Contributions of This Paper
2 The Single Elevator Problem (SEP)
2.1 Competitive Solutions
3 LA-Based Solutions
3.1 Problem Modelling
3.2 LRI-Based Solution: SEP3
3.3 PLRI-Based Solution: SEP4
4 Results and Discussions
5 Conclusions
References
Optimizing Self-organizing Lists-on-Lists Using Enhanced Object Partitioning
1 Introduction
1.1 Contributions of this Paper
1.2 Outline of this Paper
2 Theoretical Background
2.1 The Field of Learning Automata
2.2 The OMA
3 Environments with Locality of Reference
3.1 Models of Dependence
4 Adaptive Lists-on-Lists (LOL)
5 Hierarchical Data ``Sub''-Structures
6 EOMA-Augmented Hierarchical SLLs-on-SLLs
7 Results and Discussions
8 Conclusion
References
EduBAI: An Educational Platform for Logic-Based Reasoning
1 Introduction
2 Platform Design and Methodology
2.1 Main Concepts, Rules and Game Mechanics
2.2 Axiomatizing the Game Dynamics
2.3 Axiomatizing the Team Tactic Intelligence
3 Implementation
4 Related Work and Conclusions
References
Machine Learning - Natural Language
A Machine Learning Tool for Interpreting Differences in Cognition Using Brain Features
1 Introduction
2 Methods
2.1 Dataset and Participants
2.2 Data and Image Processing
2.3 Factor Analysis
2.4 Cross-Validation Procedure
2.5 Feature Importance Analysis
3 Results
4 Conclusion
References
Comparison of the Best Parameter Settings in the Creation and Comparison of Feature Vectors in Distributional Semantic Models Across Multiple Languages
1 Introduction
2 Background
3 Data and Evaluation Methods
4 Our Heuristic Analysis
5 Results
5.1 Results of the First Phase
5.2 Results of the Second Phase
5.3 Results on the Test Datasets
6 Evaluation and Discussion
7 Conclusions
References
Distributed Community Prediction for Social Graphs Based on Louvain Algorithm
Abstract
1 Introduction
2 Related Work
2.1 Divisive Algorithms
2.2 Agglomerative Algorithms
2.3 Transformation Methods
2.4 Other Approaches
3 Proposed Methodology
3.1 Graph Statistical Analysis and Feature Enrichment
3.2 Representative Subgraph Extraction
3.3 Community Prediction Model Training and Application
4 Experiments
5 Conclusions
Acknowledgment
References
Iliou Machine Learning Data Preprocessing Method for Suicide Prediction from Family History
Abstract
1 Introduction
2 Related Work
3 Data Preprocessing
4 Iliou Preprocessing Method
5 Experimental Results
6 Conclusions
References
Ontology Population Framework of MAGNETO for Instantiating Heterogeneous Forensic Data Modalities
Abstract
1 Introduction
2 Information Extraction Methods
2.1 Multimedia Content Indexing
2.2 Text Mining
2.3 Automatic Language Translation
3 MAGNETO Knowledge Base
3.1 The Underlying Ontology
3.2 Population of the MAGNETO Ontology
4 Conclusions and Future Work
References
Random Forest Surrogate Models to Support Design Space Exploration in Aerospace Use-Case
1 Introduction
2 Related Work
3 An Approach to Explore Design Space Using Random Forests Surrogate Models
3.1 Use-Case: A Design Study of Turbine Rear Structure
3.2 Surrogate Model Generation
3.3 Improve Random Forests Model Performance
3.4 Parameters Importance and Rules from RF Surrogate Models
3.5 Decision Support
4 Experimental Design
4.1 Dataset Descriptions
4.2 Hyperparameters and Configuration Selection
4.3 Experiment
5 Results and Analysis
5.1 Parameters Importance and Rules Extraction
6 Conclusions and Future Work
References
Stacking Strong Ensembles of Classifiers
1 Introduction
2 Well-Known Methods for Creating Ensembles
3 Proposed Hybrid Scheme
4 Comparisons and Experimental Results
5 Conclusions
References
Multi Agent - IoT
An Agent-Based Framework for Complex Networks
Abstract
1 Introduction
2 Related Work
3 ACONA Framework Model
3.1 Functions
3.2 Communication and Data Structures
3.3 Configuration
3.4 Complex Structures: Controller
4 Application and Results
4.1 Cognitive Architecture in Building Automation
4.2 Self-evolving Agents in a Multi-agent System
4.3 Infrastructure in an Industry 4.0 Application
5 Conclusion and Future Work
Acknowledgment
References
Studying Emotions at Work Using Agent-Based Modeling and Simulation
1 Introduction
2 Emotional Agent-Based Models of Human Behavior in the Workplace
3 Agent-Based Model of Employees Emotions
3.1 Agent Mental States
3.2 Computational Representation of the OCC Model Components and Their Corresponding Relationships
3.3 Emotion Generation Process
4 Simulation and Results Analysis
4.1 Emotions-Evoking Stimuli
4.2 Stimuli Emotional Content
5 Discussion and Conclusion
References
Towards an Adaption and Personalisation Solution Based on Multi Agent System Applied on Serious Games
Abstract
1 Introduction
2 The Game Model Cycle
3 Towards an Improved Game Model
4 System Design Architecture
4.1 The Learner Agent
4.2 The Adaption and Personalization Model
4.3 Architecture Design for Adaption and Personalization Engine
5 Conclusion and Future Work
References
Nature Inspired Flight and Robot Control - Machine Vision
Constant Angular Velocity Regulation for Visually Guided Terrain Following
1 Introduction
2 Methods
2.1 Angular Velocity Decoding Model
2.2 Control Scheme for Automatic Terrain Following
2.3 Parameter Setting
3 Experiments and Results
3.1 Moving Grating Experiments
3.2 Terrain Following Simulations
4 Conclusion and Discussion
References
Motion Segmentation Based on Structure-Texture Decomposition and Improved Three Frame Differencing
1 Introduction
2 Related Work
3 Proposed Method
3.1 Structure-Texture Image Decomposition
4 Implementation
4.1 Extraction of Frames from Video Data and Conversion to Gray-Scale
4.2 Structure-Texture Decomposition
4.3 Improved Three-Frame Differencing
4.4 Divide in Non-overlapping Blocks
4.5 Detection of Moving Objects
4.6 Post-processing
5 Experimental Results and Analysis
5.1 Qualitative Evaluation
5.2 Quantitative Evaluation
6 Conclusion
References
Using Shallow Neural Network Fitting Technique to Improve Calibration Accuracy of Modeless Robots
Abstract
1 Introduction
2 Fuzzy Error Interpolation Method
2.1 Overview of the Fuzzy Interpolation System
2.2 Membership Functions
2.3 Control Rules
2.4 Fuzzy Inference System
3 Simulation Results
4 Conclusion and Summary
References
Recommendation Systems
Banner Personalization for e-Commerce
1 Introduction
2 Related Work
3 Proposed Evaluation Metrics
4 Proposed Framework
5 Evaluation
6 Conclusion
References
Hybrid Data Set Optimization in Recommender Systems Using Fuzzy T-Norms
Abstract
1 Introduction
2 Related Work
2.1 Reproduction of Experiments in Recommender Systems Evaluation Based on Explanations
2.2 Reproducibility of Recommender Systems Experiments Based on Explanations
3 Proposed Extension and Algorithm
3.1 Hybrid Hamacher and Einstein Products
4 Experiments and Results
4.1 Evaluation Metrics
4.2 Experimental Results
4.3 Discussion
5 Conclusions and Future Work
References
MuSIF: A Product Recommendation System Based on Multi-source Implicit Feedback
Abstract
1 Introduction
2 Related Work
3 Method
3.1 Implicit Matrix Factorization (MF)
3.2 Association Rule Mining (ARM)
3.3 Initializing User Item Factor Vectors
4 Evaluation
4.1 Single-Source Evaluation with Transaction Events
4.2 Multi-source Evaluation with Transactions, Views and Addtocart Events
4.3 Single-Source vs Multi-source Comparison
4.4 Coverage
5 Conclusions and Future Work
Acknowledgments
References
On the Invariance of the SELU Activation Function on Algorithm and Hyperparameter Selection in Neural Network Recommenders
1 Introduction
2 Activation Functions
2.1 ReLU
2.2 ELU
2.3 SELU
3 Neural Network Architectures
3.1 Single Layer Feedforward Neural Network
3.2 Single Layer Feedforward Neural Network with Global Mean
3.3 Double Layer Feedforward Neural Network
4 Experimental Results
5 Conclusion
References
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken