Artificial Neural Networks and Machine Learning - ICANN 2011

Name: Artificial Neural Networks and Machine Learning - ICANN 2011 | 21st International Conference on Artificial Neural Networks, Espoo, Finland, June 14-17, 2011, Proceedings, Part I
Brand: Springer
Price: 53.49 EUR
Availability: OnlineOnly

21st International Conference on Artificial Neural Networks, Espoo, Finland, June 14-17, 2011, Proceedings, Part I

Timo Honkela Wlodzislaw Duch Mark Girolami Samuel Kaski(Herausgeber*in)

Springer (Verlag)

Erschienen am 13. Juni 2011

XXI, 390 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-642-21735-7 (ISBN)

53,49 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Title Page
Preface
Organization
Table of Contents
Transformation Equivariant Boltzmann Machines
Introduction
Building in Transformation Equivariance
Rotation Equivariant RBMs
Rotation and Translation Equivariant RBMs
Rotation and Translation Equivariant Deep Belief Nets
Inference and Learning in the Models
Experiments
Related Work
Discussion
References
Improved Learning of Gaussian-Bernoulli Restricted Boltzmann Machines
Introduction
Gaussian-Bernoulli RBM
Improved Learning of Gaussian-Bernoulli RBM
New Parameterization of the Energy Function
Parallel Tempering
Adaptive Learning Rate
Experiments
Learning Faces
Learning Natural Images
Discussion
References
A Hierarchical Generative Model of Recurrent Object-Based Attention in the Visual Cortex
Introduction
Setup
Data Sets and Plain DBM vs. RRF-DBM
Relation to Attentional Theories
Experiments: Inspection of the Hidden States
Experiments: Quantitative Evaluation
Top-Down Suppression on Sparse Representations
Spatial vs. Object-Based Attention
Conclusion
References
l_1-Penalized Linear Mixed-Effects Models for BCI
Introduction
Statistical Model
Model Setup
l_1-penalized Maximum Likelihood Estimator
Prediction of the Random-Effects
Model Selection
Computational Implementation
Available Data and Experiments
Generation of the Ensemble
Validation
Results
Subject-to-Subject Transfer
Session-to-Session Transfer
Relation of Baseline Misclassification to s2 and t2
Discussion and Conclusions
References
Slow Feature Analysis - A Tool for Extraction of Discriminating Event-Related Potentials in Brain-Computer Interfaces
Introduction
Methods
Data
Decomposition Methods
Component Selection, Feature Extraction and Classification
Results
Discussion
References
Transforming Auto-Encoders
Introduction
Learning the First Level of Capsules
More Complex 2-D Transformations
Modeling Changes in 3-D Viewpoint
Discussion
References
Stacked Convolutional Auto-Encoders for Hierarchical Feature Extraction
Introduction
Preliminaries
Auto-Encoder
Denoising Auto-Encoder
Convolutional Neural Networks
Convolutional Auto-Encoder (CAE)
Max-Pooling
Stacked Convolutional Auto-Encoders (CAES)
Experiments
Initializing a CNN with Trained CAES Weights
Conclusion
References
Error-Backpropagation in Networks of Fractionally Predictive Spiking Neurons
Introduction
Fractionally Predictive Spiking Neurons
Learning in Networks of Fractionally Predictive Spiking Neurons
Experiments
Discussion
References
ESN Intrinsic Plasticity versus Reservoir Stability
Introduction
Problem Statement
Echo State Networks as a Special Structure RNN
Intrinsic Plasticity Adaptation and Its Relation to ESN Reservoir Stability
Theoretical Stability Conditions for ESN
Numerical Investigations of ESN stability
Conclusions
References
Adaptive Routing Strategies for Large Scale Spiking Neural Network Hardware Implementations
Introduction
Adaptive NoC Router Architecture
Experiments and Results
Traffic throughput
Router Performance, Power and Area
Performance of FPGA-Based Adaptive NoC Router
Conclusion
References
Self-Organizing Map for the Multi-Goal Path Planning with Polygonal Goals
Introduction
Self-Organizing Map for Multi-Goal Path Planning with Polygonal Goals
Experiments
Conclusion
References
Unlearning in the BCM Learning Rule for Plastic Self-organization in a Multi-modal Architecture
Introduction
Model
General Architecture
Perceptive Map
Unlearning
Motivation
Equations
Properties
Results
Conclusion
References
Neuronal Projections Can Be Sharpened by a Biologically Plausible Learning Mechanis
Introduction
Methods
Results
Discussion
References
Explicit Class Structure by Weighted Cooperative Learning
Introduction
Theory and Computational Methods
Cooperative Learning
Weighted Cooperation
Controlling Cooperation and Computational Procedures
Results and Discussion
Conclusion
References
Unsupervized Data-Driven Partitioning of Multiclass Problems
Introduction
The Unsupervised Partitioning Method for Multiclass Problems
Hierarchy Construction
Classification of New Datapoints
Experimental Results
Conclusions
References
Bounds for Approximate Solutions of Fredholm Integral Equations Using Kernel Networks
Introduction
Approximation from a Dictionary
Approximate Solutions to Fredholm Integral Equations
Bounds on Approximation Errors for Dictionaries Generated by Kernels of Integral Equations
Bounds on Approximation Errors for Dictionaries Generated by Resolvent Kernels
Conclusions
References
An Improved Training Algorithm for the Linear Ranking Support Vector Machine
Introduction
Learning Setting
Algorithm Description
Computational Experiments
Conclusion
References
Extending Tree Kernels with Topological Information
Introduction
Kernels for Trees
Injecting Positional Information into Tree Kernels
Algorithmic Issues
Experiments
Conclusion and Future Work
References
Accelerating Kernel Neural Gas
Introduction
Preliminaries
Neural Gas Algorithm
Kernelized Neural Gas
Nyström Approximation of the Kernel Matrix
Sparse Coefficient Matrix
Active Learning
Experiments
Conclusions
References
State Prediction: A Constructive Method to Program Recurrent Neural Networks
Introduction
State Prediction: A Constructive Approach
Sampling Dynamics for State Prediction
Programming the Dynamics of a Single Neuron
Programming Two Neuron Circuits
Programming Input-Driven Network Dynamics
Storing Sequences in a Large Network
Conclusion
References
Cluster Self-organization of Known and Unknown Environmental Sounds Using Recurrent Neural Network
Introduction
Environmental Sounds Classification System
Multiple Timescale Recurrent Neural Network (MTRNN)
Environmental Sound Classification System
Experiments
Condition
Result
Discussion
Conclusions
References
Time-Dependent Series Variance Estimation via Recurrent Neural Networks
Introduction
Nonlinear Dynamic GARCH Modelling
The GARCH(p,q) Model
The RNN-GARCH(p,q) Model
Dynamic Training of RNN-GARCH
Experiments in Volatility Inference
Conclusion
References
Historical Consistent Complex Valued Recurrent Neural Network
Introduction
Historical Consistent Complex Valued RNN
Complex Valued Back-Propagation
Architecture Description and Insights on Training
Problem Description and Modeling Results
Conclusions and Outlook
References
Sparse Spatio-temporal Gaussian Processes with General Likelihoods
Introduction
Model and Methods
Spatio-temporal Gaussian Processes
Making the Model Tractable
Sparse Approximations
Expectation Propagation for Dynamic Systems
Results
The Effect of Sparse Approximations
Tropical Rainforest Data
Conclusions
References
Learning Curves for Gaussian Processes via Numerical Cubature Integration
Introduction
Recursion for Learning Curve
Eigenfunction Expansion Approximation of Recursion
Numerical Cubature Approximation of Recursion
Numerical Comparison
Conclusion
References
Cross-Species Translation of Multi-way Biomarkers
Introduction
Previous Work
Model
Dimensionality Reduction and Covariate Effects
Alignment of Irregular Time Series
Estimation of Shared and Specific Covariate Effects
Matching
Experiments
Generated Data
Biological Data
Conclusion
References
An Evaluation of the Image Recognition Method Using Pulse Coupled Neural Network
Introduction
The Image Recognition Using PCNN
The Pulse Coupled Neural Network Model
The Pattern Recognition Using PCNN Icons
Parameter Optimization Using Real Code Genetic Algorithm
Simulation Results
Conclusion
References
Using the Leader Algorithm with Support Vector .Machines for Large Data Sets
Introduction
Support Vector Machines
The Leader Algorithm as a Preprocessing Procedure for SMVs
The Leader Algorithm
The Proposed Approach
Experiments
Data Sets
Experimental Setting
Results
Conclusions and Future Work
References
Automatic Seizure Detection Incorporating Structural Information
Introduction
Materials and Methods
EEG Data
Feature Extraction
Classification Approaches
Results
Simulation on Randomized Training and Test Set
Real-life Setting
Computational Costs
Discussion
References
The Grouped Author-Topic Model for Unsupervised Entity Resolution
Introduction
Previous Work
Grouped Author-Topic Model
Inference
Experiments
Discussion
References
Kullback-Leibler Divergence for Nonnegative Matrix Factorization
Introduction
NMF Based on I-Divergence
NMF Based on KL-Divergence
Normalized Kullback-Leibler Divergence
Equivalence to pLSI
Projected Gradient Algorithms for NMF
Experiments
Conclusions
References
Distributed Deterministic Temporal Information Propagated by Feedforward Neural Networks
Introduction
Methods
Spiking Neuron Model
Neural Network
Dynamical Systems for Generation of the Input Spike Trains
Simulated Input Spike Trains
Pattern Detection and Reconstruction of Time Series
Similarity between Two Spike Trains
Results
Discussion
References
Chaotic Complex-Valued Multidirectional Associative Memory with Variable Scaling Factor
Introduction
Chaotic Complex-Valued Multidirectional Associative Memory with Variable Scaling Factor
Structure
Learning Process
Recall Process
External Input
Computer Experiment Results
One-to-Many Associations of Multi-valued Patterns (S=4)
Comparison of One-to-Many Associations Ability
Conclusion
References
Predicting Reaction Times in Word Recognition by Unsupervised Learning of Morphology
Introduction
Experimental Setup
Reaction Time Data and Model Evaluation
Statistics and Computational Models
Data for Learning Computational Models
Results
Discussion
References
An Examination of the Dynamic Interaction within Metaphor Understanding Using a Model Simulation
Introduction
Experiment Using Visual Primes
Method
Results of the Experiment
The Model of the Metaphor Understanding
Architecture of the Model
Model Simulation
Estimated Parameters
Discussion
References
Visual Pathways for Shape Abstraction
Introduction
Related Work
The Skeleton Extraction Process with Embedded Curvature Information
Experimental Results
References
Improving Articulatory Feature and Phoneme Recognition Using Multitask Learning
Introduction
Articulatory Feature Estimation
Previous Work
Proposed Work
Experimental Setup
Results
Articulatory Feature Classification
Phoneme Recognition
Discussion and Conclusions
References
OrBEAGLE: Integrating Orthography into a Holographic Model of the Lexicon
Introduction
Methodology
Holographic Reduced Representations
A Holographic Encoding for Word-Form
BEAGLE
Experiment
Discussion
References
On the Problem of Finding the Least Number of Features by L1-Norm Minimisation
Introduction
Feature Selection by Zero-Norm Minimisation
Experiments
Optimality of the Support Feature Machine
Preliminaries
Optimality Condition
Arguments for the Superior Results of the SFM
Conclusions
References
Extracting Coactivated Features from Multiple Data Sets
Introduction
Extraction of Coactivated Features
Modeling the Coupling between the Data Sets
Two Data Sets: A Generalization of Canonical Correlation Analysis
Analysis of Multiple Data Sets
Simulations with Artificial Data
Simulations with Real Data
Simulations with Natural Images
Simulations with Brain Imaging Data
Conclusions
References
Single Layer Complex Valued Neural Network with Entropic Cost Function
Introduction
Complex Valued NN
One Layer CVNN
Batch Learning
MEE for Learning in Batch Mode
Experiments
Datasets
Results
Conclusions
References
Batch Intrinsic Plasticity for Extreme Learning Machines
Introduction
Extreme Learning Machine
Supervised Read-Out Learning by Ridge Regression
Batch Intrinsic Plasticity
Results
Batch Intrinsic Plasticity and Single Neuron Behavior
Robotics Regression Task
Abalone Regression Task
Conclusion
References
An Empirical Study on the Performance of Spectral Manifold Learning Techniques
Introduction
Techniques
Synthetic Data Sets
Quality Measures
Experimental Results
Discussion
References
Semi-supervised Learning for WLAN Positioning
Introduction
The Semi-supervised Approach
Isomap
Manifold-Based Radio Map Learning
Calibrating the Manifold to Geographical Coordinates
Positioning
Deployment and Results
Conclusion
References
Ensemble-Teacher Learning through a Perceptron Rule with a Margin
Introduction
Model
Theory of Ensemble-Teacher Learning with a Perceptron Rule
Theory of Proposed Method
Results
Conclusion
References
Topic-Dependent Document Ranking: Citation Network Analysis by Analogy to Memory Retrieval in the Brain
Introduction
Methods
Citation Network and Spreading Activation
Algorithm: Initial-State-Dependent Retrieval of Information
Expressing a Topic by Seed Documents
Comparison Experiment: Evaluation of the Performance
Bibliographic Data
Topic-Dependent Ranking
Visualization
Results
Discussion
References
PADDLE: Proximal Algorithm for Dual Dictionaries LEarning
Introduction
Proximal Methods for Learning Dual Dictionaries
Experiments
Conclusion
References
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken