Adaptive and Natural Computing Algorithms

Name: Adaptive and Natural Computing Algorithms | 10th International Conference, ICANNGA 2011, Ljubljana, Slovenia, April 14-16, 2011, Proceedings, Part I
Brand: Springer
Price: 53.49 EUR
Availability: OnlineOnly

10th International Conference, ICANNGA 2011, Ljubljana, Slovenia, April 14-16, 2011, Proceedings, Part I

Andrej Dobnikar Uros Lotric Branko Ster(Herausgeber*in)

Springer (Verlag)

Erschienen am 9. April 2011

XVI, 433 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-642-20282-7 (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 - Part I
Plenary Session
Autonomous Discovery of Abstract Concepts by a Robot
Introduction
Discovery of Abstract Notions with ILP
Experimental Loop
Discovering Abstract Concepts by Predicate Invention in ILP
Experimental Results
Conclusions
References
Neural Networks
Kernel Networks with Fixed and Variable Widths
Introduction
Radial and Kernel Models
Minimization of Error Functionals
Scaling of Kernels
Gaussian Kernels
References
Evaluating Reliability of Single Classifications of Neural Networks
Introduction
Related Work
Reliability Estimation
Local Modeling of Prediction Error
Local Cross-Validation Reliability Estimator
Variance of a Bagged Model
Density-Based Reliability Estimator
Results
Testing Methodology
Typical Cases
Main Results
Conclusions
References
Nonlinear Predictive Control Based on Multivariable Neural Wiener Models
Introduction
Model Predictive Control Algorithms
Multivariable Neural Wiener Models
Model Predictive Control Based on Multivariable Neural Wiener Models
Quadratic Programming MPC Optimisation Problem
Implementation Details
Algorithm Summary
Simulation Results
Conclusions
References
Methods of Integration of Ensemble of Neural Predictors of Time Series - Comparative Analysis
Introduction
The Theoretical Basis of Integration of Predictors
The Averaging
Principal Component Analysis
Blind Source Separation
Application of PCA and Additional Neural Network as an Integrator
Forecasting the 24-Hour Load Pattern in Power System
The Results of Individual Predictors
The Results of Application of an Ensemble
Forecasting PM10 Pollution for the Next Day
Conclusions
References
A Rejection Option for the Multilayer Perceptron Using Hyperplanes
Introduction
Multilayer Perceptron
Rejection Option
Rejection Option Using Hyperplanes
Hyperplanes Calculation
Experiments and Results
Conclusions
References
Parallelization of Algorithms with Recurrent Neural Networks
Introduction
First Translation: Programs to Networks
Second Translation: Networks to Tuples
Results and Discussion
Conclusions
References
Parallel Training of Artificial Neural Networks Using Multithreaded and Multicore CPUs
Introduction
Multilayer Perceptrone Neural Network
Backpropagation Learning
Parallelized Neural Network Training
Parallel Backpropagation Learning
Multithreaded Training Implementation
Simulations and Results
Datasets Used for Simulations
Efficiency of Parallelization
Conclusions
References
Supporting Diagnostics of Coronary Artery Disease with Neural Networks
Introduction
Methods
Stepwise Diagnostic Process
Image Classification with Neural Networks
ArTex and ARes Algorithms for Multi-resolution Image Parameterization
Materials
Scintigraphic Images
Results
Results in CAD Diagnostics
Explaining the Neural Network's Predictions by Evaluation of Individual Attribute's Contributions
Discussion
References
The Right Delay Detecting Specific Spike Patterns with STDP and Axonal Conduction Delays
Introduction
Spike-Timing-Dependent Plasticity
Materials and Methods
The Neuron Model
The Network
STDP Implementation
Detecting Specific Spatio-Temporal Spike Patterns
Discussion
References
New Measure of Boolean Factor Analysis Quality
Introduction
Generative Model of Boolean Factor Analysis
Information Gain
Probabilities Estimation
Properties of Information Gain
Summary
References
Mechanisms of Adaptive Spatial Integration in a Neural Model of Cortical Motion Processing
Introduction
Neural Model
Method for Motion Discontinuity Detection
Model of Contextual Enhancement
Model of Diffusion Process
Combination of Methods
Results
Moving Box
"Yosemite with Clouds" Sequence
Discussion and Conclusions
References
Self-organized Short-Term Memory Mechanism in Spiking Neural Network
Introduction
Stabilizing Spiking Neuron (SS Neuron)
Synaptic Plasticity
Network
Spatio-temporal Patterns and Short-Term Memory
Experiments
Results
Conclusion
References
Approximation of Functions by Multivariable Hermite Basis: A Hybrid Method
Introduction
Approximation Framework
One-Variable Hermite Functions
Multivariable Function Approximation
Multivariable Hermite Basis
Scaling Parameters
Basis Selection
Examples
Conclusions
References
Using Pattern Recognition to Predict Driver Intent
Introduction
Benefit of Gaze as a Data Source
Manoeuvre Recognition
Patterns of Gaze Behaviour
Simulator Study
Driving Task
Equipment
Data Processing
Artificial Neural Networks
Analysis and Results
Conclusion
References
Neural Networks Committee for Improvement of Metal's Mechanical Properties Estimates
Introduction
Problem Statement and Data Analysis
Neural Networks Committee
Train Procedure
Stability of the Committee Decision
Conclusion
References
Logarithmic Multiplier in Hardware Implementation of Neural Networks
Introduction
Iterative Logarithmic Multiplier
Mathematical Formulation
Hardware Implementation
Multilayer Perceptron with Highly Parallel Neural Unit
Experimental Work
Conclusion
References
Efficiently Explaining Decisions of Probabilistic RBF Classification Networks
Introduction
Explanation Methods
Method EXPLAIN
Method IME
Probabilistic RBF Networks
Marginalization of PRBF
Instance Level and Model Level Explanation of PRBF
Conclusions
References
Evolving Sum and Composite Kernel Functions for Regularization Networks
Introduction
Regularization Networks
Sum and Composite Kernel Functions
Genetic Parameter Search
Experiments
Conclusion
References
Optimisation of Concentrating Solar Thermal Power Plants with Neural Networks
Introduction
The Simulation of CSP Plants
The Energetic Model of a CSP Plant
The Economic Model of a CSP Plant
Use of Genetic Algorithms to Optimise Solar Plants
Improvements of the Optimisation
Bilinear Interpolation
Artificial Neural Networks
Experimental Results
Conclusion and Outlook
References
Emergence of Attention Focus in a Biologically-Based Bidirectionally-Connected Hierarchical Network
Introduction
Our Approach
A Closer Look at the Network Used in the Simulations
Data Set
Procedure
Results and Analysis
Multiple Objects without Any Attention
Multiple Objects with Bottom-Up Attentional Effects
Multiple Objects with Both Bottom-Up as Well as Top-Down Attentional Effect
Conclusions
References
Visualizing Multidimensional Data through Multilayer Perceptron Maps
Introduction
Multidimensional Data Visualization
Data Mapping through Multilayer Perceptrons
Results
Conclusion
References
Input Separability in Living Liquid State Machines
Introduction
Motivation
Neuronal Cell Cultures
Data Acquisition and Stimulation System
Living Reservoir LSM Experiment
Method Details
Classification of LNN Responses to Sensory Input
Stimulus Artifact Suppression Techniques
Response Classification and Determination of Separable Pattern Subsets
Classifier Parameter Optimization
Genetic Algorithm-Based Approach for Finding Best Pattern Subsets
Results
Conclusions and Future Work
Closed-Loop Input Pattern Separability Optimization
Characteristics of Separable Patterns
References
Predictive Control of a Distillation Column Using a Control-Oriented Neural Model
Introduction
Model Predictive Control Algorithms
Classical Linear MPC of a Distillation Column
Nonlinear MPC Using Control-Oriented Neural Model
Control-Oriented Neural Model
Neural Models and Training
Nonlinear MPC of a Distillation Column
Conclusions
References
Neural Prediction of Product Quality Based on Pilot Paper Machine Process Measurements
Introduction
Related Work
Pilot Paper Machine Experiments
Method for Predicting Quality from Process State
Trial Point Selection and Preprocessing
Neural Model
Results
Discussion
References
A Robotic Scenario for Programmable Fixed-Weight Neural Networks Exhibiting Multiple Behaviors
Multiple Behaviors in a Fixed-Weight Neural Network
A Programmable Fixed-Weight CTRNN
A Multiplicative CTRNN
A Programmable Neural Network for Robot Control
The Robotic Scenario
Discussion
References
Self-Organising Maps in Document Classification: A Comparison with Six Machine Learning Methods
Introduction
The Data Sets
Data Set 1: Reuters-21578, Mod Apte Split
Data Set 2: 20 Newsgroups, Matlab/Octave
Data Set 3: Spanish CLEF 2003
Preprocessing
Document Classification with Self-Organising Map
The Baseline Methods
Results
Conclusions and Discussion
References
Analysis and Short-Term Forecasting of Highway Traffic Flow in Slovenia
Introduction
Traffic Flow Data
Traffic Flow Data Analysis
Forecasting Framework
Benchmark Models
ARIMA and Neural Network Models
Forecasting Results
Conclusions
References
Evolutionary Computation
A New Method of EEG Classification for BCI with Feature Extraction Based on Higher Order Statistics of Wavelet Components and Selection with Genetic Algorithms
Introduction
Dataset Description
Proposed Feature Extraction Method Based on DWT and HOS
Proposed Feature Selection Method Based on Genetic Algorithms
Conclusion
References
Regressor Survival Rate Estimation for Enhanced Crossover Configuration
Introduction
Related Work
Regressor Based Hierarchical Representation
Fuzzy Controlled Encapsulation
Variable Size and Shape Schemata
Application
Conclusions
References
A Study on Population's Diversity for Dynamic Environments
Introduction
Evolutionary Algorithms
Genetic Operators
Dynamic Test Environments
Experimental Study
Experimental Design
Results on Cyclic Environments
Results on Random Environments
Conclusions
References
Effect of the Block Occupancy in GPGPU over the Performance of Particle Swarm Algorithm
Introduction
Related Work
Methods and Materials
GPGPU Cards
ParallelModels of Evolutionary Algorithm
Particle Swarm Optimizer
Production Setup
Analysis and Results
Conclusion and Future Work
References
Two Improvement Strategies for Logistic Dynamic Particle Swarm Optimization
Introduction
Logistic Dynamic Particle Swarm Optimization
Canonical PSO
Logistic Dynamic PSO
Two Improvement Strategies on LDPSO
Mutation Strategy Based Improvement
Selection Strategy Based Improvement
Experiment and Analysis
Experiment Setting
Comparison between LDPSO and Canonical PSO
Performance of Mutation Strategy
Performance of Selection Strategy
Conclusion
References
Digital Watermarking Enhancement Using Wavelet Filter Parametrization
Introduction
Digital Watermarking in the DWT Domain
Genetic-Based Digital Image Watermarking Enhancement
Embedding Algorithm
Wavelet Parametrization
Wavelet Synthesis Algorithm
Experimental Results
Conclusions
References
CellularDE: A Cellular Based Differential Evolution for Dynamic Optimization Problems
Introduction
Differential Evolution
The Proposed Algorithm
Memory of the Cells
Cellular Differential Evolution
Maintaining Diversity of Population
Dealing with the Change
Experiments
Moving Peaks Benchmark
Experimental Settings
Experimental Results
Conclusion
References
Optimization of Topological Active Nets with Differential Evolution
Introduction and Previous Work
Brief Description of Topological Active Nets
Differential Evolution
Results
Comparison with a Genetic Algorithm
Hybridization of Differential Evolution and Greedy Search
Conclusions
References
Study on the Effects of Pseudorandom Generation Quality on the Performance of Differential Evolution
Introduction
Pseudorandom Number Generation in Evolutionary Methods
Empirical Framework
Numerical Results and Analysis
Conclusions and Further Work
References
Sensitiveness of Evolutionary Algorithms to the Random Number Generator
Introduction
Related Work
Random Number Generator
The Random Number Generators Tested
Evolutionary Algorithms
Statistical Inference
Non-parametric Statistical Inference
Production Setup
Results and Analysis
Conclusions and Future Work
References
New Efficient Techniques for Dynamic Detection of Likely Invariants
Introduction
Daikon Algorithm
Problem Space
Daikon Disadvantage
New Ideas for Improving Invariant Detection
The First Idea
The Second Idea
Comparison of the Results
Conclusion and Further Works
References
Classification Ensemble by Genetic Algorithms
Introduction
Background
Classifier Ensemble
Genetic Algorithm
Regression
Related Works
Dynamic Estimation of the Best Local Classifier
Estimation of Competence Regions
Proposed Method
Problem Definition
Challenges of the Problem
Genetic Algorithm Based Classifier Ensemble
Experimental Results
References Datasets
Parameters Setting
Conclusion
References
Simulated Evolution (SimE) Based Embedded System Synthesis Algorithm for Electric Circuit Units (ECUs)
Introduction
Previous Work
Proposed Algorithm
SimE for the ECU Synthesis Problem
Scheduling Algorithms
Fault Tolerance (FT) Schemes
Simulations
Conclusion
References
Taxi Pick-Ups Route Optimization Using Genetic Algorithms
Introduction
Asymmetric Travel Salesman Problem
Mathematical Formulation of the Problem
Methodology
GA Parameters
Fitness Function
Implementation
Experimental Results
Conclusion and Future Work
References
Optimization of Gaussian Process Models with Evolutionary Algorithms
Introduction
Modeling with Gaussian Processes
Evolutionary Algorithms
Genetic Algorithm
Differential Evolution
Particle Swarm Optimization
Experiments and Results
Conclusion
References
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken