
Artificial Intelligence XXXVI
Description
Alles über E-Books | Antworten auf Fragen rund um E-Books, Kopierschutz und Dateiformate finden Sie in unserem Info- & Hilfebereich.
This book constitutes the proceedings of the 39th SGAI International Conference on Innovative Techniques and Applications of Artificial Intelligence, AI 2019, held in Cambridge, UK, in December 2019.
The 29 full papers and 14 short papers presented in this volume were carefully reviewed and selected from 49 submissions. The volume includes technical papers presenting new and innovative developments in the field as well as application papers presenting innovative applications of AI techniques in a number of subject domains. The papers are organized in the following topical sections: machine learning; knowledge discovery and data mining; agents, knowledge acquisition and ontologies; medical applications; applications of evolutionary algorithms; machine learning for time series data; applications of machine learning; and knowledge acquisition.More details
Other editions
Additional editions

Persons
Content
- Intro
- Preface
- Organization
- Contents
- Technical Papers
- CascadeML: An Automatic Neural Network Architecture Evolution and Training Algorithm for Multi-label Classification (Best Technical Paper)
- 1 Introduction
- 2 Related Work
- 2.1 Neural Networks for Multi-label Classification and BPMLL
- 2.2 The Cascade2 Algorithm
- 3 The CascadeML Algorithm
- 4 Experiment Design
- 5 Results
- 6 Conclusions and Future Work
- References
- Machine Learning, Knowledge Discovery and Data Mining
- Purity Filtering: An Instance Selection Method for Support Vector Machines
- 1 Introduction
- 2 Relevant Previous Work
- 3 Purity Filtering
- 3.1 Data Reduction
- 3.2 Hyperplane Estimation
- 3.3 Margin Estimation
- 3.4 Data Selection
- 3.5 Theoretical Analysis
- 4 Experimentation and Discussion
- 4.1 Synthetic Data
- 4.2 Covertype Dataset
- 5 Conclusions
- References
- Towards Model-Based Reinforcement Learning for Industry-Near Environments
- 1 Introduction
- 2 Literature Review
- 2.1 Automated Storage and Retrieval Systems (ASRS)
- 2.2 Model-Based Reinforcement Learning
- 3 Background
- 4 Learning Policies Using Predictive Models
- 4.1 Motivation and Environment Safety
- 4.2 The Dreaming Variational Autoencoder V2
- 4.3 Model Selection
- 4.4 Implementation
- 5 The Deep Warehouse Environment
- 5.1 Motivation
- 5.2 Implementation
- 6 Experimental Results
- 6.1 The Importance of Compute
- 6.2 Results
- 7 Conclusion and Future Work
- References
- Stepwise Evolutionary Learning Using Deep Learned Guidance Functions
- 1 Introduction
- 2 Deep Learned Guidance Functions
- 2.1 Formalisation
- 3 Example: Applying Deep LGFs to the Rubik's Cube
- 3.1 Constructing the LGF
- 3.2 Sources of Error
- 4 Experiments and Results
- 5 Related Work
- 6 Summary and Future Work
- References
- Monotonicity Detection and Enforcement in Longitudinal Classification
- 1 Introduction
- 2 Background
- 2.1 Longitudinal Classification
- 2.2 Monotonic Classification
- 2.3 Monotonicity Measures
- 2.4 The XGBoost Tree Boosting Algorithm
- 3 Methodology
- 3.1 Quantitative Measures for Monotonicity Detection
- 3.2 The Proposed Longitudinal Monotonicity Detection Approaches
- 3.3 Datasets
- 3.4 Experimental Setup
- 4 Results of the Experiments and Comparative Analysis
- 4.1 Predictive Accuracies of Constructed Models
- 4.2 Monotonicity Constraints and Their Effects on Model Sizes
- 4.3 Feature Importance
- 5 Conclusion
- References
- Understanding Structure of Concurrent Actions
- 1 Introduction
- 2 Related Work
- 3 Preliminaries
- 4 Implicit Structure: Action Elimination Using Task Invariance
- 5 Explicit Structure: Clustering of Action Space
- 6 Experiments
- 6.1 Implicit Structure - Action Elimination
- 6.2 Explicit Structure - Clustering
- 7 Conclusion and Future Work
- References
- Agents, Knowledge Acquisition and Ontologies
- Demonstrating the Distinctions Between Persuasion and Deliberation Dialogues
- 1 Introduction
- 2 Speech Acts
- 2.1 Refinement of Speech Acts
- 3 Design
- 3.1 Agent Class
- 3.2 Dialogue Class
- 3.3 PersuasionDialogue Class
- 3.4 DeliberationDialogue Class
- 3.5 Simulation Class
- 4 Realisation
- 4.1 Project Architecture
- 4.2 Interface
- 5 Description of the Tool
- 6 Testing and Evaluation
- 6.1 Unit Testing
- 7 Discussion and Concluding Remarks
- References
- Ontology-Driven, Adaptive, Medical Questionnaires for Patients with Mild Learning Disabilities
- Abstract
- 1 Introduction
- 2 Background and Related Work
- 3 Methodology
- 4 Ontologies Development
- 4.1 Medical Questionnaire Ontology
- 4.2 Accessibility Ontology
- 5 System Implementation - Java Adaptive Engine
- 5.1 Java Engine
- 5.2 Dynamic Changes to Stack
- 6 Scenario Based Evaluation
- 7 Conclusion and Future Work
- References
- Exposing Knowledge: Providing a Real-Time View of the Domain Under Study for Students
- 1 Introduction
- 2 Current Solutions
- 3 Proposed Knowledge Search Framework
- 3.1 Background Worker
- 3.2 User Interface
- 3.3 Similarity Analysis
- 4 Evaluation
- 5 Conclusion
- References
- Short Technical Papers
- A General Approach to Exploit Model Predictive Control for Guiding Automated Planning Search in Hybrid Domains
- 1 Introduction
- 2 Background
- 3 Using MPC to Guide Automated Planning Search
- 4 Experimental Analysis
- 5 Conclusion
- References
- A Tsetlin Machine with Multigranular Clauses
- 1 Introduction
- 2 A Tsetlin Machine with Multigranular Clauses
- 3 Experimental Results
- 4 Conclusion
- References
- Building Knowledge Intensive Architectures for Heterogeneous NLP Workflows
- Abstract
- 1 Introduction
- 2 Related Work
- 3 An Architecture for Knowledge Intensive Workflows
- 4 DeepKAF Evaluation
- 5 Conclusions
- References
- WVD: A New Synthetic Dataset for Video-Based Violence Detection
- 1 Introduction
- 2 Hot vs Cold Weapons Violence Dataset
- 3 Experimental Evaluation
- 4 Conclusion and Future Work
- References
- Application Papers
- Evolving Prediction Models with Genetic Algorithm to Forecast Vehicle Volume in a Service Station (Best Application Paper)
- Abstract
- 1 Introduction
- 2 Background
- 3 Model Formulation
- 3.1 Neural Network Based Prediction Model
- 3.2 Genetic Algorithm
- 4 Experiment Setup and Analysis of Results
- 4.1 Comparison of Accuracy Using MAPE
- 4.2 Comparison of Produced Parameter Settings
- 4.3 Typical GA Evolution of Model
- 5 Conclusion
- References
- Medical Applications
- Are You in Pain? Predicting Pain and Stiffness from Wearable Sensor Activity Data
- 1 Introduction
- 2 Related Work
- 3 Participants and Data Acquisition
- 3.1 Participants and Recruitment Procedure
- 3.2 Data Collection
- 4 Transformation of Actigraph Data
- 4.1 Generating 60s Epoch Measures
- 5 Generating Actigraphy Variables from 60s Epoch Measures
- 5.1 Intra-DAD Measures
- 5.2 Inter-DAD Measures
- 6 Predicting Morning Pain and Stiffness
- 6.1 Feature Selection and Modelling Algorithms
- 6.2 Evaluation Method
- 6.3 Results
- 7 Conclusions and Future Work
- References
- Motif Discovery in Long Time Series: Classifying Phonocardiograms
- 1 Introduction
- 2 Previous Work
- 3 PCG Frequent Motif Selection and Extraction
- 3.1 Cycle Segmentation
- 3.2 Candidate Motif Selection
- 3.3 Frequent Motif Extraction
- 4 Evaluation
- 4.1 Evaluation Data
- 4.2 Experimental Set-Up
- 4.3 Cycle Segmentation Subprocess Evaluation
- 4.4 Candidate Motif Selection Subprocess Evaluation
- 4.5 Frequent Motif Extraction Subprocess Evaluation
- 4.6 Runtime Evaluation
- 4.7 Classification Accuracy
- 5 Conclusions
- References
- Exploring the Automatisation of Animal Health Surveillance Through Natural Language Processing
- Abstract
- 1 Introduction
- 1.1 Study Scope: Salmonellosis (S. Dublin) and Pneumonia NOS
- 2 Materials and Method
- 2.1 Materials
- 2.2 Lexical Analysis of APHA 'Surveillance Highlights' Free-Text Datasets
- 2.3 NLP Tasks: Named-Entity Recognition and Similarity and Relatedness
- 2.4 Assessing the Performance for the NLP Tasks
- 3 Results
- 3.1 Lexical Analysis of APHA Surveillance Highlights Free-Text Datasets
- 3.2 NLP Tasks: Name-Entity Recognition and Similarity and Relatedness
- 3.3 Assessing the Performance of the NLP Tasks
- 4 Discussion
- 5 Conclusion
- Acknowledgements
- References
- Applications of Evolutionary Algorithms
- GenMuse: An Evolutionary Creativity Enhancement Tool
- Abstract
- 1 Introduction
- 2 The System
- 3 The Methodologies
- 3.1 Genetic Algorithms
- 3.2 Interactive Genetic Algorithms (IGAs)
- 3.3 Neural Networks
- 4 Practical Use of the Two Methodologies - GAs and ANNs
- 4.1 The Genetic Algorithm Used for This Project
- 4.2 Riff Creation
- 4.2.1 Encoding Chromosomes
- 4.2.2 Evaluating the Fitness Value of a Solution
- 4.3 The Artificial Neural Network
- 4.3.1 Interpretation of Classification
- 4.3.2 Training the Artificial Neural Network
- 5 Results
- 6 Conclusions
- 7 Future Work
- References
- Evolutionary Art with an EEG Fitness Function
- Abstract
- 1 Background
- 1.1 Context
- 1.2 Interactive Art Projects
- 1.3 Related Work
- 1.4 Genetic Algorithms
- 1.5 Fitness Function
- 1.6 Roulette Wheel Selection
- 1.7 Mutation and Crossover
- 2 Methodology
- 2.1 Analysis of Mondrian's Paintings
- 2.2 Physical Construction
- 3 Testing
- 4 Discussion
- 5 Conclusion
- References
- A Multi-objective Design of In-Building Distributed Antenna System Using Evolutionary Algorithms
- Abstract
- 1 Introduction
- 2 Related Work
- 3 Problem Formulation
- 4 Optimization Model
- 4.1 A Solution Representation for IB-DAS
- 4.2 NSGA II Approach to Solving the Problem
- 5 Experimental Results
- 5.1 Pareto-Optimal Front Representation
- 5.2 Comparison of NSGA II Vs. GA
- 5.3 Run Time Analysis of NSGA II
- 6 Conclusion and Future Work
- References
- Machine Learning for Time Series Data
- Investigation of Machine Learning Techniques in Forecasting of Blood Pressure Time Series Data
- 1 Introduction
- 2 Forecasting Strategies
- 3 Machine Learning Algorithms
- 4 Methodology
- 4.1 Data Set
- 4.2 Forecast Model Formulation
- 4.3 Results
- 5 Discussion
- 6 Conclusions
- References
- Stock Index Forecasting Using Time Series Decomposition-Based and Machine Learning Models
- 1 Introduction
- 2 Non-classical Decomposition Models
- 2.1 Discrete Wavelet Transform
- 2.2 Empirical Mode Decomposition
- 2.3 Variational Mode Decomposition
- 3 Ensemble Forecasting Framework
- 3.1 Data
- 3.2 Methodology
- 4 Results and Discussion
- 4.1 Performance Comparison of Ensemble Models with Standalone Models
- 4.2 Comparison of Performance Among Decomposition-Based ANN Models and SVR Models
- 5 Threats to Validity
- 6 Conclusions
- References
- Effective Sub-Sequence-Based Dynamic Time Warping
- 1 Introduction
- 2 Background
- 3 Previous Work
- 4 Sub-Sequence-Based DTW
- 5 Time Complexity
- 6 Evaluation
- 6.1 Data Sets
- 6.2 Evaluation Results
- 7 Conclusion
- References
- Applications of Machine Learning
- Developing a Catalogue of Explainability Methods to Support Expert and Non-expert Users
- 1 Introduction
- 2 Related Work
- 3 Use-Case - Explaining Engineering Notes to Desk-Based Agents
- 3.1 The Dataset
- 3.2 Evaluation of Classification Engine
- 3.3 Results and Discussion
- 3.4 Co-creation of Explanation Methods
- 4 The Explanation Engine
- 4.1 Low-Level Explanations
- 4.2 High-Level Explanations
- 4.3 Co-creation of Explanations
- 5 The Application
- 6 Evaluation of Explanation Framework
- 6.1 Results and Discussion
- 7 Conclusions
- References
- A Generic Model for End State Prediction of Business Processes Towards Target Compliance
- 1 Introduction
- 2 A Framework of End State Prediction of Business Process Data
- 3 Dataset Description
- 3.1 Pre-processing
- 3.2 Feature Extraction
- 4 Results and Discussion
- 5 Conclusion
- References
- An Investigation of the Accuracy of Real Time Speech Emotion Recognition
- Abstract
- 1 Introduction
- 2 Proposed System
- 2.1 Hardware and Programming Language
- 2.2 Voice Detector
- 2.3 Pre-processing
- 2.4 Feature Extraction
- 2.5 Feature Normalisation
- 2.6 Pattern Recognition Algorithms
- 3 Experiments
- 3.1 The Dataset
- 3.2 Experiments on Speech Databases
- 3.3 Real Time Speech Emotion Case-Based Study
- 4 Evaluation
- 4.1 Results Evaluation
- 4.2 Performance on the Raspberry Pi 3 B+
- 4.3 Comparison with Other Speech Emotion Systems
- 5 Conclusions and Future Work
- References
- Contributing Features-Based Schemes for Software Defect Prediction
- Abstract
- 1 Introduction
- 2 Related Work
- 3 Contributing Features-Based Logistic Regression
- 3.1 A Case Study of Contributing Feature Selection
- 4 The Prediction Models
- 5 Experiments and Analysis
- 5.1 Results and Discussions
- 5.2 Comparative Analysis
- 6 Conclusion
- Acknowledgement
- References
- Induction Motor Inter-turn Short Circuit Fault Detection Using Efficient Feature Extraction for Machine Learning Based Fault Detectors
- Abstract
- 1 Introduction
- 2 Literature Survey
- 3 Features Extracted from the GOTIX Vibration Dataset
- 4 Features Extracted from the GOTIX Electric Dataset
- 5 Features from the Envelope Spectrums of the Vibration Signals
- 6 Features from the Average Spectrum Versus Order for Vibration Signals
- 7 Conclusion
- References
- Hybrid Feature Selection Method for Improving File Fragment Classification
- 1 Introduction
- 2 Related Work
- 3 Proposed Hybrid Feature Selection Method
- 3.1 Data Collection/Extraction
- 3.2 File Fragmentation
- 3.3 Feature Extraction
- 3.4 Features Selection
- 3.5 Classification
- 4 Experiments and Results
- 4.1 Data and Feature Extraction
- 4.2 Experiments for Selecting Appropriate Filters and Classifiers
- 4.3 Implementing Hybrid Feature Selection Method
- 4.4 File Fragment Classification Results
- 5 Conclusion
- References
- Optimization of Silicon Tandem Solar Cells Using Artificial Neural Networks
- Abstract
- 1 Introduction
- 2 Previous Works
- 3 Dataset Description
- 4 The Architecture of an Artificial Neural Network
- 5 Results and Discussions
- 6 Conclusion
- References
- Stochastic Local Search Based Feature Selection for Intrusion Detection
- Abstract
- 1 Introduction
- 2 Background
- 2.1 Classification
- 2.2 Feature Selection
- 2.3 Stochastic Local Search
- 3 Proposed Approach
- 3.1 Proposed Architecture for Intrusion Detection
- 3.2 Solution Representation
- 3.3 SLS for Feature Selection
- 3.4 SLS with Classifier for Intrusion Detection
- 3.5 The Considered Classifiers
- 4 Experiments
- 4.1 The Considered Datasets
- 4.2 Evaluation Measures
- 4.3 Parameters Tuning and Impacts
- 4.4 Numerical Results
- 4.5 Further Comparison
- 5 Conclusion
- References
- Knowledge Acquisition
- Improving the Adaptation Process for a New Smart Home User
- 1 Introduction
- 2 Related Work
- 3 User-Guided Transfer Learning (UTL)
- 3.1 Survey
- 3.2 Simulation
- 3.3 Activity Recognition
- 3.4 Transfer Learning
- 4 User-Guided Transfer Learning (UTL) Interface
- 5 Validation
- 6 Discussion
- 7 Conclusion
- References
- Short Application Papers
- Analysis of Electronic Health Records to Identify the Patient's Treatment Lines: Challenges and Opportunities
- Abstract
- 1 Introduction
- 2 Challenge for Reconstructing the Patient's Treatment Lines
- 2.1 Named Entity Recognition Challenges
- 2.2 Temporal Relation Identification Challenges
- 2.3 The Integration of Structured Results Challenges
- 3 Solution
- 4 Conclusion and Future Work
- Acknowledgment
- References
- Characterisation of VBM Algorithms for Processing of Medical MRI Images
- 1 Introduction
- 1.1 Image Registration (Co-registration) Algorithms
- 2 Methodology
- 2.1 Temporal Cortex
- 3 Discussion
- References
- Analogical News Angles from Text Similarity
- 1 Introduction
- 2 Assumptions
- 3 Finding Unmatched News Reports
- 4 Similarity Measures
- 5 Results
- 6 Conclusion and Further Work
- References
- Mindfulness Mirror
- Abstract
- 1 Introduction
- 2 Genetic Algorithms
- 3 Interactive Genetic Algorithms
- 4 Mindfulness
- 5 Bio Feedback and Human Issues
- 6 Construction of the System
- 7 Results and Discussion
- 8 Conclusion
- References
- Predicting Bid Success with a Case-Based Reasoning Approach: Explainability and Business Acceptance
- Abstract
- 1 Background
- 2 Case-Based Reasoning: An Alternative
- 3 Case-Based Reasoning Approach
- 4 User Engagement and Acceptance
- 5 Conclusions
- References
- Data Augmentation for Ambulatory EEG Based Cognitive State Taxonomy System with RNN-LSTM
- 1 Introduction
- 2 Proposed Work
- 2.1 Data Collection Procedure
- 2.2 Proposed Methodology
- 2.3 Data Augmentation Method
- 3 Result Analysis and Discussion
- 4 Conclusion and Future Work
- References
- Time-Series-Based Classification of Financial Forecasting Discrepancies
- 1 Introduction
- 2 Background
- 3 Dataset Description
- 4 Methods
- 5 Experiments
- 6 Conclusion
- References
- Predicting Soil pH by Using Nearest Fields
- 1 Introduction
- 2 Related Work
- 3 Predicting Soil pH
- 3.1 Soil Dataset
- 3.2 Features Based on Nearest Fields
- 3.3 Data Mining Techniques for Prediction
- 4 Experimental Results
- 5 Conclusion and Future Work
- References
- Information Retrieval for Evidence-Based Policy Making Applied to Lifelong Learning
- 1 Introduction
- 2 The Policy Data Model
- 2.1 Descriptive Feature Model
- 2.2 Policy Retrieval Model
- 3 Query Models for Policy Search
- 3.1 Free-Text Queries
- 3.2 Free-Text Structured Queries
- 3.3 Constrained Structured Queries
- 4 Evaluation
- 5 The ENLIVEN Dataset
- 6 Background and Related Works
- 7 Conclusion and Research Directions
- References
- On Selection of Optimal Classifiers
- Abstract
- 1 Introduction
- 2 Research Background
- 2.1 Machine Learning
- 2.2 Pareto Optimality
- 3 Research Methodology
- 4 Experiments and Results
- 4.1 The Dataset
- 4.2 Results
- 5 Conclusion and Future Work
- References
- Author Index
System requirements
File format: PDF
Copy protection: Watermark-DRM (Digital Rights Management)
System requirements:
- Computer (Windows; MacOS X; Linux): Use the free software Adobe Reader, Adobe Digital Editions, or any other PDF viewer of your choice (see eBook Help).
- Tablet/Smartphone (Android; iOS): Install the free app Adobe Digital Editions or another reading app for eBooks, e.g., PocketBook (see eBook Help).
- E-reader: Bookeen, Kobo, Pocketbook, Sony, Tolino and many more (only limited: Kindle).
The file format PDF always displays a book page identically on any hardware. This makes PDF suitable for complex layouts such as those used in textbooks and reference books (images, tables, columns, footnotes). Unfortunately, on the small screens of e-readers or smartphones, PDFs are rather annoying, requiring too much scrolling.
This eBook uses Watermark-DRM, a „soft” copy protection. This means that there are no technical restrictions to prevent illegal distribution. However, there is a personalised watermark embedded in the eBook that can be used to identify the purchaser of the eBook in the event of misuse and to provide evidence for legal purposes.
For more information, see our eBook Help page.