Advances in Artificial Intelligence

Name: Advances in Artificial Intelligence | 33rd Canadian Conference on Artificial Intelligence, Canadian AI 2020, Ottawa, ON, Canada, May 13-15, 2020, Proceedings
Brand: Springer
Price: 53.49 EUR
Availability: OnlineOnly

33rd Canadian Conference on Artificial Intelligence, Canadian AI 2020, Ottawa, ON, Canada, May 13-15, 2020, Proceedings

Cyril Goutte Xiaodan Zhu(Editor)

Springer (Publisher)

Published on 5. May 2020

XV, 572 pages

E-Book

PDF with digital watermarking

System requirements

978-3-030-47358-7 (ISBN)

€53.49incl. 7% vat

System requirements

for PDF with digital watermarking

E-Book Single Licence

Available for download

Description

More details

Other editions

Content

Intro
Preface
Organization
Contents
Toward Adversarial Robustness by Diversity in an Ensemble of Specialized Deep Neural Networks
1 Introduction
2 Specialists Ensemble
2.1 Ensemble Construction
2.2 Voting Mechanism
3 Analysis of Specialists Ensemble
4 Experimentation
4.1 Empirical Results
5 Related Works
6 Conclusion
References
Locating Influential Agents in Social Networks: Budget-Constrained Seed Set Selection
1 Introduction
2 Methods
2.1 Graph Decomposition Methods
2.2 Information Diffusion Models
3 Results
3.1 Datasets
3.2 Subgraph Decomposition Properties
3.3 Analyzing Trust
3.4 Evaluating Spreading Performance
4 Discussion
5 Conclusions and Future Extensions
References
Investigating Relational Recurrent Neural Networks with Variable Length Memory Pointer
1 Introduction
2 Related Work
3 The Model
4 Experimental Setup
5 Experimental Results and Analysis
6 Conclusion and Future Work
References
Unsupervised Monocular Depth Estimation CNN Robust to Training Data Diversity
1 Introduction
2 Related Work
3 Method
3.1 Depth Estimation as Image Reconstruction
3.2 Input Image Correction
3.3 Camera Parameters Estimation
3.4 Total Training Loss
4 Results
4.1 Implementation Details
4.2 Datasets
4.3 Depth Estimation Performance Evaluation
5 Conclusion
References
The K-Closest Resemblance Classifier for Remote Sensing Data
1 Introduction
2 The K-Closest Resemblance Classifier
3 Experiments and Evaluation
4 Results and Discussion
5 Conclusion
References
Reinforcement Learning in a Physics-Inspired Semi-Markov Environment
1 Introduction
1.1 Contributions
2 Related Work
3 Semi-Markov Phase Change Environment
4 Experimental Setup
4.1 Reinforcement Learning Algorithms
4.2 Evaluation Method
5 Results
5.1 DQN on the Semi-Markov Phase Change Environment
5.2 DRQN on the Semi-Markov Phase Change Environment
5.3 Agents with Hindsight Experience Replay
6 Discussion
7 Conclusion
References
Deep Multi Agent Reinforcement Learning for Autonomous Driving
1 Introduction
2 Related Work
3 Background on Multi-Agent Reinforcement Learning
4 Methods
4.1 Multi-Agent Message Sharing Network (MA-MeSN)
4.2 Multi-Agent Broadcast Network (MA-BoN)
5 Experimental Methodology
5.1 Treadmill Driving Environment
6 Results and Discussions
6.1 Centralized Training on Multi-Agent Driving Environment
6.2 Ablation Study of Scalability of MA-BoN
6.3 Theoretical Study of Emergent Communication
6.4 Fully Decentralized Cooperative Policy in Driving Environment
7 Conclusion and Future Work
References
Incremental Sequential Rule Mining with Streaming Input Traces
1 Introduction
2 Background
2.1 Literature Review
2.2 Sequential Pattern Mining
2.3 Sequential Rule Mining
3 ERMiner
4 Incremental Sequential Rule Mining
4.1 Information Maintenance
4.2 The IERMiner Algorithm
4.3 IERMiner Example
5 Experiments
5.1 Experiment Akosarak 5000
5.2 Experiment Akosarak 25k
5.3 Experiment MSNBC 31790
6 Concluding Remarks
References
FASTT: Team Formation Using Fair Division
1 Introduction
2 Related Work
2.1 Team Formation
2.2 Fair Division
3 A Framework for Fair Division with Dependent Values
4 Algorithms
5 Empirical Evaluation
5.1 Results
6 Conclusions and Future Work
References
Empirical Confidence Models for Supervised Machine Learning
1 Introduction
2 Related Works
3 Empirical Confidence Models
3.1 Reliability Measures as Meta-Features
3.2 Meta Training Data
4 Numerical Results
5 Conclusion
References
Selection Driven Query Focused Abstractive Document Summarization
1 Introduction
2 Proposed Model
2.1 Task Description
2.2 Embedding Layer
2.3 Input Concatenation Layer
2.4 Encoding Layer
2.5 InceptionNet Layer
2.6 Self-attention Layer
2.7 Gated Layer
2.8 Attention Layer
2.9 Final State Concatenation Layer
2.10 Output Embedding Layer
2.11 Decoding Layer
2.12 Output Layer
3 Experiments and Results
4 Conclusion
References
VecHGrad for Solving Accurately Tensor Decomposition
1 Motivation
2 Related Work
3 Background
3.1 Tensor Operations
3.2 Tensor Decomposition
4 VecHGrad for Tensor Decomposition
4.1 Introduction to VecHGrad for Vectors
4.2 VecHGrad for Accurate Resolution of Tensor Decomposition
5 Experiments
6 Conclusion
References
Sensitivity to Risk Profiles of Users When Developing AI Systems
1 Introduction and Background
2 Desiderata for Trusted AI
3 Reasoning with Differing User Preferences for Trusted AI
3.1 Game-Theoretic Reasoning
3.2 Distinguishing Trust and Risk Averseness
4 Fairness and Explainability
5 Acquiring and Updating User Risk Profiles
6 Conclusion and Future Work
References
Forecasting Seat Counts in the 2019 Canadian Federal Election Using Twitter
1 Introduction
2 Twitter Data
3 Riding Prediction Model
3.1 Global Vote Ratio Forecasting
3.2 Riding Distribution Model
4 Training and Testing the Model
5 Conclusion and Future Work
References
Adapting Ensemble Neural Networks to Clinical Prediction in High-Dimensional Settings
1 Introduction
2 Probabilistic Survival Model
2.1 Neural Network Architecture Proposal
2.2 Classical Approaches
3 Experiment
3.1 Simulated Dataset
3.2 Models
4 Conclusion
References
A Cost Skew Aware Predictive System for Chest Drain Management
1 Introduction
2 Previous Work
3 Cost Matrix Genetic Algorithm Optimization
4 Domain Description and Pre-processing Methods
5 Experimental Framework
6 Conclusion
References
Topological Data Analysis for Arrhythmia Detection Through Modular Neural Networks
1 Introduction
2 Topological Data Analysis
3 Deep-Learning Approach
3.1 Datasets
3.2 Preprocessing
3.3 Auto-Encoder
3.4 Architecture
4 Experimental Results
4.1 Channel Comparison
4.2 Arrhythmia Detection
4.3 Arrhythmia Classification
5 Benchmarks Comparison
5.1 Premature Ventricular Heartbeats Detection
5.2 8-Classes Classification
6 Conclusion
References
Big Players: Emotion in Twitter Communities Tweeting About Global Warming
1 Introduction
2 Related Work
3 Methodology
4 Results and Analysis
5 Discussion
6 Conclusion
References
Using Topic Modelling to Improve Prediction of Financial Report Commentary Classes
1 Introduction
2 Methodology
2.1 Topic Modelling
2.2 Learning Model
3 Results
4 Conclusions and Future Work
References
Wise Sliding Window Segmentation: A Classification-Aided Approach for Trajectory Segmentation
1 Introduction
2 Definitions
3 Related Works
4 The Proposed Method
4.1 Generating the Error Signal
4.2 Creating Training Data
4.3 Binary Classification Model
4.4 Majority Vote
5 Experimental Evaluation
5.1 Datasets
5.2 Experimental Setup
5.3 Results and Discussion
6 Conclusions
References
Using Deep Reinforcement Learning Methods for Autonomous Vessels in 2D Environments
1 Introduction
2 Related Works
3 A New DRL Method for Unmmaned Surface Vessels
3.1 Definitions
3.2 Baseline and Our Proposed Method
4 Experiments
4.1 Dataset Creation and Evaluation Metrics
4.2 Training and Testing Setup
4.3 Result Analysis and Discussion
5 Conclusions and Future Works
References
CB-DBSCAN: A Novel Clustering Algorithm for Adjacent Clusters with Different Densities
1 Introduction
2 Preliminaries
2.1 DBSCAN
2.2 Multi-density Clustering
3 Related Work
4 CB-DBSCAN
4.1 Mini-Clustering
4.2 Centroid-Based Merging
5 Clustering Evaluation
6 Time Complexity Analysis
7 Conclusion and Future Work
References
Anomaly Detection and Prototype Selection Using Polyhedron Curvature
1 Introduction
2 Background on Polyhedron Curvature
3 Anomaly Detection
3.1 Curvature Anomaly Detection
3.2 Kernel Curvature Anomaly Detection
3.3 Anomaly Landscape and Anomaly Paths
4 Prototype Selection
4.1 Inverse Curvature Anomaly Detection
4.2 Kernel Inverse Curvature Anomaly Detection
5 Experiments
5.1 Experiments for Anomaly Detection
5.2 Experiments for Prototype Selection
6 Conclusion and Future Direction
References
Ethical Requirements for AI Systems
1 Introduction
2 Ethical Principles and Codes
3 Deriving Ethical Requirements
4 The Case of Driveless Cars
5 Related Work
6 Conclusions
References
A Deep Neural Network for Counting Vessels in Sonar Signals
1 Introduction
2 Materials
3 Proposed Methods
3.1 Vessel Detection
3.2 Counting Targets
4 Experiments
4.1 Vessel Detection
4.2 Counting Targets
5 Conclusion
References
Partial Label Learning by Entropy Minimization
1 Introduction
2 Related Work
3 Proposed Method
4 Experiment
5 Conclusion
References
Low-Dimensional Dynamics of Encoding and Learning in Recurrent Neural Networks
1 Introduction
2 Methods
2.1 Experimental Setup
2.2 Experimental Datasets
2.3 The Spectrum of Lyapunov Exponents
3 Results
3.1 Development of a Task Relevant Structure
3.2 Formation of Limit Cycles
4 Summary and Conclusions
References
From Explicit to Implicit Entity Linking: A Learn to Rank Framework
1 Introduction
2 Proposed Approach
2.1 Term-Based and String-Based Features
2.2 Graph-Based Features
3 Datasets and Experimental Setup
4 Results and Discussion
References
Automatic Polyp Segmentation Using Convolutional Neural Networks
1 Introduction
1.1 Motivation and Contributions
2 Related Work
3 Methods and Materials
3.1 Experimental Dataset
3.2 Data Pre-processing
3.3 Feature Extraction Using Transfer Learning Strategy
3.4 Ensemble Method
3.5 Evaluation Criteria
4 Experiments and Results
4.1 Experimental Setup
4.2 Results and Discussion
5 Conclusion
References
Augmented Out-of-Sample Comparison Method for Time Series Forecasting Techniques
1 Introduction
2 Background
2.1 Time-Series Prediction
2.2 Time-Series Model Comparison Techniques
3 Augmented Out-of-Sample Comparison Technique
4 Experiments
4.1 Experimental Settings
4.2 Benefits of Augmented Out-of-Sample Testing
4.3 Results
5 Conclusion
References
Predicting the Number of Reported Bugs in a Software Repository
1 Introduction
2 Methodology
2.1 Data
2.2 ACF and PACF
2.3 Forecasting Models
3 Results
4 Threats to Validity
4.1 Construct Validity
4.2 Internal Validity
4.3 External Validity
5 Conclusion and Future Work
References
Evaluation of a Failure Prediction Model for Large Scale Cloud Applications
1 Introduction
2 Related Work
3 Failure Prediction Model
4 Experiments and Evaluation Results
4.1 Trace Description
4.2 Experimental Setup
4.3 Failure Analysis and Characterization
4.4 Classifiers and Prediction Techniques
4.5 Feature Selection Algorithms
5 Conclusion and Future Work
References
Customer Segmentation and Churn Prediction in Online Retail
1 Introduction
2 Background
2.1 Analytic Hierarchy Process (AHP)
2.2 RFM Analysis
3 The Proposed Model
3.1 Data Pre-processing
3.2 Standardization
3.3 Clustering
3.4 Calculating Weights
3.5 Calculating the CLV Score for Clusters
3.6 Churn Prediction
4 Experimentation
4.1 Dataset Description
4.2 Customer Segmentation
4.3 Churn Prediction
5 Conclusions
References
Detection and Diagnosis of Breast Cancer Using a Bayesian Approach
1 Introduction
2 Related Work
3 Methods and Proposed Bayesian Network Model
4 Results
4.1 Bayesian Network
4.2 Conditional Probability Querying
4.3 Bayesian Network Classifier
5 Threats to Validity
6 Conclusion and Future Work
References
Query Focused Abstractive Summarization via Incorporating Query Relevance and Transfer Learning with Transformer Models
1 Introduction
2 Related Work
3 Our Proposed Approach
4 Debatepedia Dataset
5 Experimental Setup
5.1 Baselines
5.2 Training Parameters
6 Results and Discussions
7 Conclusions and Future Work
References
Word Representations, Seed Lexicons, Mapping Procedures, and Reference Lists: What Matters in Bilingual Lexicon Induction from Comparable Corpora?
1 Introduction
2 Methods
2.1 Bag of Words
2.2 Embedding Methods
2.3 Cross-domain Similarity Local Scaling (CSLS)
3 Data and Analysis
4 Results
5 Analysis
6 Conclusion
References
Attending Knowledge Facts with BERT-like Models in Question-Answering: Disappointing Results and Some Explanations
1 Introduction
2 Related Work
3 Experimental Protocol
4 Models
4.1 Model A
4.2 Model B
4.3 Model C
4.4 Model D
5 Implementation Details
6 Experiments
7 Conclusion
References
Machine Learning the Donor Journey
1 Introduction
2 Related Research
3 Problem Formulation
4 Our Approach
4.1 Preliminary Experimental Setup
4.2 Email Optimization
5 Empirical Evaluation
5.1 Training the RNN
5.2 Full Email Experiments
6 Discussion and Future Work
References
Exploring Deep Anomaly Detection Methods Based on Capsule Net
1 Introduction
2 Insights into Existing Work
2.1 Boundary-Based Methods
2.2 Distribution-Based Methods
2.3 Benchmarks
3 CapsNet-Based Normality Score Functions
3.1 Prediction-Probability-Based Normality Score
3.2 Reconstruction-Error-Based Normality Score
3.3 Combined Normality Score
4 Experiments
4.1 Comparison on MNIST, Fashion-MNIST, and Small-Norb Data
4.2 Case Studies in Normality Score Functions
5 Conclusion
References
Question-Worthy Sentence Selection for Question Generation
1 Introduction
2 Related Work
2.1 Question Generation
2.2 Feature and Graph-Based Sentence Ranking and Selection
3 Methodology
3.1 Feature-Based Question-Worthy Sentence Extraction
3.2 Context-Aware Question Generation
4 Experimental Setup and Results
4.1 Dataset and Implementation Details
4.2 Evaluation Metrics
4.3 Question-Worthy Context Results
4.4 Question Generation Results
5 Conclusion and Future Work
References
Challenges in Vessel Behavior and Anomaly Detection: From Classical Machine Learning to Deep Learning
1 Introduction
2 Research Challenges and Opportunities
2.1 Behavior Detection: Supervised vs. Unsupervised
2.2 Identifying Recurrent Behavior Patterns
2.3 Towards Interpretable Behavior Patterns
2.4 Diving into Deep Learning
2.5 Big, Yet Limited Data
3 Summary and Final Remarks
References
An Energy-Efficient Method with Dynamic GPS Sampling Rate for Transport Mode Detection and Trip Reconstruction
1 Introduction
2 Related Works
2.1 Trip Reconstruction
2.2 Transport Modes Detection
2.3 Combined Approach
3 Model and Algorithm
4 Experimentations
4.1 Data for the Model
4.2 GPS Data Collection
5 Results
5.1 Transport Mode Detection
5.2 Trip Reconstruction
5.3 Energy Consumption
6 Conclusion
References
Similarity Matching of Temporal Event-Interval Sequences
1 Introduction
2 Problem Statement
3 Similarity Matching of Interval-Based Temporal Sequences
3.1 Matching Using Relative Frequency
3.2 Matching Using Position Code
3.3 Matching Using Multiple Kernel Learning
4 Experiments
5 Conclusion
References
Classification of Rare Recipes Requires Linguistic Features as Special Ingredients
1 Introduction
2 Related Work
3 Datasets
4 Model Design
4.1 The Neural Sub-model
4.2 The Linguistic Sub-model
4.3 The Fusion Component
4.4 Training
5 Results and Discussion
6 Conclusion
References
Happiness Analysis with Fisher Information of Dirichlet-Multinomial Mixture Model
1 Introduction
2 Problem Statement
3 The Mathematical Model
4 Experimental Evaluations
5 Conclusion
References
Personalized Multi-Faceted Trust Modeling in Social Networks
1 Introduction and Background
2 Dataset and Trust Indicators
3 Experiment Description
4 Experiments and Results
5 Discussion and Conclusion
References
Mixing ICI and CSI Models for More Efficient Probabilistic Inference
1 Introduction
2 Background
2.1 NAT Modelling of ICI
2.2 CPT-Tree Modelling of CSI
3 Orthogonality of NAT and CSI Models
4 Mixed NAT-CSI Bayesian Networks
5 Formalizing CPT-Tree Transformation
6 Inference with Mixed NAT-CSI Bayesian Networks
7 Experiments
8 Conclusion and Remarks
References
RideSafe: Detecting Sexual Harassment in Rideshares
1 Introduction
2 Relevant Work
3 Dataset and Data Preprocessing
4 Experiments and Analysis
5 Threats to Validity
6 Future Work and Conclusion
References
Amalgamated Models for Detecting Duplicate Bug Reports
1 Introduction
2 Related Work
3 Dataset and Pre-processing
3.1 Dataset
3.2 Pre-processing of Textual Features
4 Methodology
4.1 Latent Dirichlet Allocation
4.2 Term Frequency-Inverse Document Frequency
4.3 Word2Vec
4.4 Global Vectors for Word Representation
4.5 Proposed Amalgamated Model
5 Evaluation Metrics
5.1 Recall-Rate@k
5.2 Mean Average Precision (MAP)
5.3 Mean Reciprocal Rank (MRR)
6 Results and Discussion
6.1 Empirical Analysis
6.2 Effectiveness of Amalgamation of Models
6.3 Statistical Significance and Effect Size
7 Threats to Validity
8 Conclusion
References
Investigating Citation Linkage as a Sentence Similarity Measurement Task Using Deep Learning
1 Introduction
2 Citation Linkage
3 Related Work
4 Citation Linkage Corpus Creation
5 Citation Linkage as a Semantic Similarity Measurement Task
6 Experimental Setup and Analysis of the Results
6.1 Network Parameters and Settings
6.2 Performance Analysis
7 Conclusions and Future Work
References
Improving Classification Using Topic Correlation and Expectation Propagation
1 Introduction
2 Background
2.1 Latent Dirichlet Allocation
2.2 Expectation Propagation
2.3 Generalized Dirichlet Distribution
3 Related Work
4 Latent Generalized Dirichlet Allocation
4.1 Model
4.2 Inference
4.3 Parameter Estimation
5 Results
6 Conclusions
References
A Scheme for Generating a Dataset for Anomalous Activity Detection in IoT Networks
1 Introduction
2 Related Work
3 Testbed Architecture
4 Analysis
4.1 Binary Classification
4.2 Category Classification
4.3 Sub-category Classification
5 Conclusion
References
Lexical Data Augmentation for Text Classification in Deep Learning
1 Introduction
2 Design Principles of PLSDA
2.1 Substitution Candidate Selection
2.2 Instance Generation
3 Performance Evaluation
3.1 Overall Performance
3.2 Effectiveness of POS Types
3.3 Sampling Strategy
4 Conclusion
References
A Deeper Look at Bongard Problems
1 Introduction
2 Related Work
3 Our Models
4 Evaluating Bongard Problem Solvers
4.1 Measuring Validity
4.2 Measuring Robustness
4.3 Measuring Simplicity
5 Experiments and Results
5.1 Setup
5.2 Results and Observations
5.3 Rule Visualization
6 Conclusions
References
Adversarial Models for Deterministic Finite Automata
1 Introduction
2 Transition Importance of a DFA
2.1 Transition Importance Estimation as an Adversarial Model Problem
2.2 Transition Importance
2.3 Evaluation of DFA Transition Importance
3 Critical Patterns of DFA
3.1 Different Types of Critical Patterns
3.2 DFA Synchronizing Algorithm
3.3 Evaluation of DFA Pattern Identification
4 Conclusion
References
Personalized Student Attribute Inference
1 Introduction
2 Personalized Models
2.1 Personalized Student Attribute Inference (PSAI)
2.2 PSAI Algorithm
3 Experiments and Results
4 Conclusion
References
Vehicle Traffic Estimation Using Weather and Calendar Data
1 Introduction
2 Background
3 Approach
3.1 Number of New Vehicles per Image Prediction
3.2 Traffic Flow Rate and Density Prediction
4 Empirical Studies Thus Far
5 Conclusion
References
Predicting Aggressive Responsive Behaviour Among People with Dementia
1 Introduction
1.1 Problem Statement
2 Approach
2.1 Data Preparation
2.2 Recurrent Neural Network Model Development
3 Conclusion and Future Work
References
Towards Analyzing the Sentiments in the Fields of Automobiles and Real-Estates with Specific Focus on Arabic Online Reviews
1 Problem Statement and Motivation
2 Literature Review
3 Research Methodology
3.1 Data Gathering and Annotation
3.2 Data Preprocessing and Features Selection
3.3 Datasets Preparation
3.4 Data Processing and Visualization
4 The Results
5 Conclusion and Future Work
References
Author Index

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Advances in Artificial Intelligence

Description

More details

Other editions

Additional editions

Content

System requirements