Advances in Self-Organizing Maps, Learning Vector Quantization, Clustering and Data Visualization

Name: Advances in Self-Organizing Maps, Learning Vector Quantization, Clustering and Data Visualization | Proceedings of the 13th International Workshop, WSOM+ 2019, Barcelona, Spain, June 26-28, 2019
Brand: Springer
Price: 149.79 EUR
Availability: OnlineOnly

Proceedings of the 13th International Workshop, WSOM+ 2019, Barcelona, Spain, June 26-28, 2019

Alfredo Vellido Karina Gibert Cecilio Angulo José David Martín Guerrero(Editor)

Springer (Publisher)

1st Edition

Published on 27. April 2019

XII, 342 pages

E-Book

PDF with digital watermarking

System requirements

978-3-030-19642-4 (ISBN)

€149.79incl. 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
Steering Committee
Program Committee
Contents
Self-organizing Maps: Theoretical Developments
Look and Feel What and How Recurrent Self-Organizing Maps Learn
1 Introduction
2 Methods
2.1 Algorithm
2.2 Representations
2.3 Evaluation
3 Results
3.1 Ambiguous Observations
3.2 Long Term Dependencies
3.3 Adapting to a Changing Dynamics
3.4 Noisy Observations
3.5 Perturbed by a Noise State
4 Conclusion
References
Self-Organizing Mappings on the Flag Manifold
1 Introduction
2 Introduction to Flag Manifold with Data Analysis Examples
3 Numerical Representation and Geodesics
3.1 Flag Manifold
3.2 Geodesic and Distance Between Two Points on Flag Manifold
3.3 Iterative Alternating Algorithm
4 SOM on Flag Manifolds
4.1 Numerical Experiment
5 Conclusions and Future Work
References
Self-Organizing Maps with Convolutional Layers
1 Introduction
2 Self-Organizing Maps
3 Related Work
4 Convolutional Layers
5 SOM with Convolutional Layers
6 Quality Measures
6.1 Kruskal Shepard Error
6.2 Cross Entropy
6.3 Minor Class Occurrence
6.4 Class Scatter Index
7 Experimental Analysis
7.1 Experimental Settings
7.2 Quality Measure Results
7.3 Visualization Results
8 Conclusion
References
Cellular Self-Organising Maps - CSOM
1 Introduction
2 Self-Organising Maps: SOM and Cellular SOM
2.1 SOM: Self-Organising Maps
2.2 CSOM: Cellular Self-Organising Maps
2.3 Algorithms
3 Experimental Setup and Results
3.1 Quantisation of Artificial d-dimensional Distributions
3.2 Video Compression
4 Conclusion
References
A Probabilistic Method for Pruning CADJ Graphs with Applications to SOM Clustering
1 Introduction: The CADJ Graph
2 A Probabilistic Model for CADJ
3 A Multi-focal View
4 The Metric
5 Ranking Connections for Removal
6 Clustering Applications
6.1 6d Synthetic Spectral Image
6.2 Real Data: Ocean City Spectral Image
7 Conclusions and Outlook
References
Practical Applications of Self-Organizing Maps, Learning Vector Quantization and Clustering
SOM-Based Anomaly Detection and Localization for Space Subsystems
Abstract
1 Introduction
2 Related Work
3 Self-Organizing Map Background
4 Methods
4.1 Data Processing
4.2 Anomaly Detection via MQE
4.3 Anomaly Localization via Supervised Feature Extraction
5 Experiments and Discussion
5.1 Data Collection
5.2 Anomaly Detection Analysis
5.3 Anomaly Localization Analysis
6 Conclusions and Future Work
References
Self-Organizing Maps in Earth Observation Data Cubes Analysis
1 Introduction
2 Land Use and Cover Change Information from Earth Observation Data Cubes
2.1 Earth Observation Satellite Image Time Series
2.2 Vegetation Indexes
2.3 Using SOM to Improve the Quality of Land Use and Cover Samples
3 Case Study
4 Final Remarks
References
Competencies in Higher Education: A Feature Analysis with Self-Organizing Maps
Abstract
1 Introduction
2 State of the Art
3 Materials and Methods
3.1 Training Dataset
3.2 Clustering Students and Obtaining Main Features
4 Results
5 Conclusions and Future Works
References
Using SOM-Based Visualization to Analyze the Financial Performance of Consumer Discretionary Firms
Abstract
1 Introduction
2 Literature Review
3 Methodology
4 Results
5 Discussion
6 Conclusion
References
Novelty Detection with Self-Organizing Maps for Autonomous Extraction of Salient Tracking Features
1 Introduction
2 Image Representation with SOM
2.1 Self-Organizing Maps
2.2 Image Representation
3 Dynamic Neural Fields
4 Our Tracking Application
5 Results
6 Conclusion
References
Robust Adaptive SOMs Challenges in a Varied Datasets Analytics
Abstract
1 Introduction
2 SOM Algorithm
3 RA-SOM Algorithm
4 Simulation Results
4.1 Balance Dataset
4.2 Dermatology Dataset
4.3 Arcene Dataset
4.4 Gisette Dataset
5 Conclusion and Future Work
References
Detection of Abnormal Flights Using Fickle Instances in SOM Maps
1 Introduction
2 The Data
3 First Level of Labeling
4 Two-Levels Clustering and Resulting Labels
5 Dissimilarity Matrix and Relational SOM
5.1 Substitutions Costs
5.2 Adding Costs and Deletion Costs
6 Clustering the Labeled Sequences and Identifying Fickle Flights
7 Conclusion
References
LVQ-type Classifiers for Condition Monitoring of Induction Motors: A Performance Comparison
1 Introduction
2 Basics of Cluster Validation Techniques
2.1 Cluster Validity Indices
3 Prototype-Based Classifiers
3.1 LVQ Classifiers
4 Results and Discussion
5 Conclusions and Further Work
References
When Clustering the Multiscalar Fingerprint of the City Reveals Its Segregation Patterns
1 Introduction
2 Building a Multiscalar Fingerprint of the City
3 Focal Distances and Distortion Coefficients
4 Clustering Trajectories
4.1 Defining Contrasts and Indices of Features Importance
4.2 Five hotspots of Segregation for the City of Paris
5 Conclusion and Perspectives
References
Using Hierarchical Clustering to Understand Behavior of 3D Printer Sensors
Abstract
1 Introduction
1.1 3D Printing Overview
1.2 Data Collection and Parsing
1.3 Data Preprocessing
2 Statistical Clustering Method
3 Interpretation of Clusters
3.1 Analysis of Conditional Distributions Versus Classes
3.2 Detection of Non-informative and Redundant Variables
3.3 Pattern Conceptualization
4 Conclusion
References
A Walk Through Spectral Bands: Using Virtual Reality to Better Visualize Hyperspectral Data
1 Introduction
2 Background
2.1 Hyperspectral Data
2.2 Virtual Reality for Data Visualization
3 Example Visualizations
3.1 Indian Pines
3.2 Chemical Plume Detection
4 Conclusion
References
Incremental Traversability Assessment Learning Using Growing Neural Gas Algorithm
1 Introduction
2 Problem Specification
3 Evaluation Results
4 Conclusion
References
Learning Vector Quantization: Theoretical Developments
Investigation of Activation Functions for Generalized Learning Vector Quantization
1 Introduction
2 Generalized Learning Vector Quantization - A Multilayer Network Perspective
2.1 Basics of GLVQ
2.2 GLVQ - A Neural Network Perspective
2.3 Activation Function for MLP and GLVQ-MLN
3 Numerical Results for Activation Functions in GLVQ
3.1 Data Sets
3.2 Results
4 Conclusions
References
Robustness of Generalized Learning Vector Quantization Models Against Adversarial Attacks
1 Introduction
2 Learning Vector Quantization
3 Experimental Setup
3.1 Adversarial Attacks
3.2 Robustness Metrics
3.3 Training Setup and Models
4 Results
5 Conclusion
References
Passive Concept Drift Handling via Momentum Based Robust Soft Learning Vector Quantization
1 Introduction
2 Related Work
3 Streaming Data and Concept Drift
3.1 Concept Drift
4 Robust Soft Learning Vector Quantization
4.1 Momentum Based Optimization
5 Experiments
5.1 Results
6 Conclusion
References
Prototype-Based Classifiers in the Presence of Concept Drift: A Modelling Framework
1 Introduction
2 Models and Methods
2.1 Learning Vector Quantization
2.2 The Dynamics of LVQ
2.3 LVQ Dynamics Under Concept Drift
3 Results and Discussion
4 Summary and Outlook
References
Theoretical Developments in Clustering, Deep Learning and Neural Gas
Soft Subspace Topological Clustering over Evolving Data Stream
1 Introduction
2 Model Proposition
3 Experimental Evaluation
4 Conclusion
References
Solving a Tool-Based Interaction Task Using Deep Reinforcement Learning with Visual Attention
1 Introduction
2 Reinforcement Learning
2.1 The REINFORCE Algorithm
3 The RAA3C MODEL
3.1 The Location Network
3.2 Glimpse Network
3.3 Context Network
3.4 The Actor-Critic Network
3.5 Training
4 Learning Domain
5 Experiments/Results
6 Conclusion
References
Approximate Linear Dependence as a Design Method for Kernel Prototype-Based Classifiers
1 Introduction
2 Basics of Prototype-Based Classification
2.1 Kernel Functions
3 The Proposed Approach
4 Results and Discussion
4.1 Initial Tests
4.2 More General Tests
5 Conclusions and Further Work
References
Subspace Quantization on the Grassmannian
1 Introduction
2 The Grassmannian
3 Averaging Subspaces
4 Grassmann K-means Algorithm
5 The LBG Algorithm on the Grassmannian
6 Numerical Experiments
6.1 MNIST Results
6.2 Indian Pines Results
7 Conclusions
References
Variants of Fuzzy Neural Gas
1 Introduction
2 Interpretation of Distance for Different Types of Data
3 Possibilistic Fuzzy c-Means
4 Possibilistic Fuzzy Neural Gas
4.1 Vectorial Data
4.2 Relational Data
4.3 Median Data
4.4 Remarks
5 Experiments
5.1 Artificial Gaussian Distributions
5.2 Clustering Transcripts of Psychotherapy Sessions
6 Conclusion
References
Autoencoders Covering Space as a Life-Long Classifier
1 Introduction
2 Related Work
3 Analysis
4 Method
5 Results
5.1 Incremental Training of Binary Classification
5.2 Catastrophic Forgetting Evaluation on MNIST
5.3 Discussion
6 Conclusion
References
Life Science Applications
Progressive Clustering and Characterization of Increasingly Higher Dimensional Datasets with Living Self-organizing Maps
Abstract
1 Introduction
2 The Living SOM
3 Clustering Comparisons to Kohonen SOMs
4 Reproducibility Is Affected by Data Insertion Order
5 Discussion
Acknowledgments
References
A Voting Ensemble Method to Assist the Diagnosis of Prostate Cancer Using Multiparametric MRI
Abstract
1 Introduction
2 Data
3 Classification Methods
4 Model Evaluation: Using a Voting Ensemble Method for Aiding Prostate Cancer Diagnosis
5 Classification Results: Application to Prostate Lesion Findings
6 Discussion: Explanation of the Classification Results Using Generative Topographic Mapping
6.1 Borderline Cases
7 Conclusions and Further Work
Acknowledgments
References
Classifying and Grouping Mammography Images into Communities Using Fisher Information Networks to Assist the Diagnosis of Breast Cancer
Abstract
1 Introduction
2 Material
3 Methodology
4 Results and Discussion
5 Conclusion
References
Network Community Cluster-Based Analysis for the Identification of Potential Leukemia Drug Targets
1 Introduction
2 Materials and Methods
2.1 Graph-Theoretical Centralities
2.2 Random Graph Null Hypothesis: The Erdös-Rényi model
2.3 Community Finding Algorithms
2.4 Data Gathering and Building the Gene-Interaction Network
3 Experimental Results and Discussion
3.1 Description of the Network Structure
3.2 Community Analysis
3.3 Analysis of the Network Based on Centrality Measures
4 Conclusions
References
Searching for the Origins of Life - Detecting RNA Life Signatures Using Learning Vector Quantization
1 Introduction
2 Biological and Artificial RNA Sequence Data - The Search for Signatures of Life
3 Methods and Similarity Measures for RNA/DNA Sequence Comparison
4 Generalized Learning Vector Quantization for Classification Learning
4.1 Stochastic Gradient Descent Learning in GLVQ
4.2 Median and Relational GLVQ
5 Results
6 Conclusion
References
Simultaneous Display of Front and Back Sides of Spherical SOM for Health Data Analysis
Abstract
1 Introduction
2 Analysis for Relationship Between Components
3 Analysis of Health Data at Prefectural Level
4 Summary
Acknowledgement
References
Author Index

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Advances in Self-Organizing Maps, Learning Vector Quantization, Clustering and Data Visualization

Description

More details

Other editions

Additional editions

Content

System requirements