Engineering Applications of Neural Networks

€ 111,39
Bisher € 111,89
pdf eBook
Sofort lieferbar (Download)
August 2009



This book constitutes the refereed proceedings of the 11th International Conference on Engineering Applications of Neural Networks, EANN 2009, held in London, GB, on August 27-29, 2009.The 47 revised full papers were carefully reviewed and selected from many submissions. The following diverse topics were discussed: reducing urban concentration, entropy topography in epileptic electroencephalography, phytoplanktonic species recognition, revealing the structure of childhood abdominal pain data, robot control, discriminating angry and happy facial expressions, flood forecasting and assessing credit worthiness.


1;Preface;5 2;Organization;6 3;Table of Contents;8 4;Intelligent Agents Networks Employing Hybrid Reasoning: Application in Air Quality Monitoring and Improvement;13 4.1;Introduction;13 4.1.1;Aim of This Project;13 4.1.2;Air Pollution;14 4.2;Theoretical Background;14 4.2.1;Risk Evaluation Using Fuzzy Logic;15 4.3;The Agent-Based System;16 4.3.1;Systems Architecture;16 4.3.2;Sensor Agents Architecture;17 4.3.3;Evaluation Agents;19 4.3.4;Systems Agents;21 4.3.5;Decision Agents;23 4.3.6;Actuators;24 4.4;Graphical User Interface;24 4.5;Pilot Application of the System;25 4.5.1;Running the Agent Network for Heat Index;25 4.5.2;Running the Agent Network for Air Pollutants;26 4.6;Conclusions;26 4.7;References;27 5;Neural Network Based Damage Detection of Dynamically Loaded Structures;29 5.1;Introduction;29 5.2;Methodology of Damage Detection;30 5.2.1;Artificial Neural Network;32 5.2.2;Stochastic Analysis;32 5.3;Software Tools;33 5.4;Application Cantilever Beam;34 5.5;Conclusions;37 5.6;References;37 6;Reconstruction of Cross-Sectional Missing Data Using Neural Networks;40 6.1;Introduction;40 6.2;Developing a New Algorithm;41 6.2.1;The Modified GRNN (Generalized Regression Neural Networks) Algorithm;42 6.2.2;The GMI Algorithm;42 6.3;Assessing the New Technique;44 6.4;Results and Discussion;45 6.5;Summary and Conclusion;45 6.6;References;46 7;Municipal Creditworthiness Modelling by Kernel-Based Approaches with Supervised and Semi-supervised Learning;47 7.1;Introduction;47 7.2;Municipal Creditworthiness Problem Description;48 7.3;Basic Notions of Support Vector Machines and Learning;49 7.4;Modelling and Analysis of the Results;52 7.4.1;Modelling by SVMs with Supervised Learning;52 7.4.2;Modelling by Kernel-Based Approaches with Semi-supervised Learning;53 7.5;Conclusion;55 7.6;References;56 8;Clustering of Pressure Fluctuation Data Using Self-Organizing Map;57 8.1;Introduction;57 8.2;Acquisition of Pressure Fluctuation Data;58 8.3;Methodology for Clustering by the SOM;59 8.3.1
;Batch Self-Organizing Map;59 8.3.2;Procedures of Clustering for Classification of Pressure Fluctuation Data and Operational Conditions;60 8.4;Results and Discussions;61 8.4.1;Simulations of Clustering of Operational Conditions;61 8.4.2;Prediction of Dynamic Behavior of Interface Based on Clustering Map;64 8.5;Conclusions;65 8.6;References;66 9;Intelligent Fuzzy Reasoning for Flood Risk Estimation in River Evros;67 9.1;Introduction;67 9.1.1;Necessity for a New Approach;68 9.1.2;Necessity for Applying Flexible Models;69 9.2;The Fuzzy Algebra Model;70 9.2.1;Implementation of the IS;73 9.3;Application in the Case of the Flood Risk Problem;73 9.4;Discussion Comparison to Existing Approaches;75 9.5;References;77 10;Fuzzy Logic and Artificial Neural Networks for Advanced Authentication Using Soft Biometric Data;79 10.1;Introduction;79 10.2;System Architecture;80 10.3;Audio Hard Feature Extraction and Authentication;81 10.4;Fingerprint Hard Feature Extraction and Authentication;83 10.5;Soft-Biometric Feature Extraction;84 10.6;Artificial Neural Network-Based Soft-Biometric Feature Scoring;85 10.7;Fuzzy Logic-Based Fusion and Authentication;85 10.8;Performance Evaluation;88 10.9;Embedded Implementation;88 10.10;Conclusions;89 10.11;References;89 11;Study of Alpha Peak Fitting by Techniques Based on Neural Networks;91 11.1;Introduction;91 11.1.1;Existing Solutions;92 11.1.2;Proposed Solution;93 11.2;Method;93 11.2.1;Training Data;93 11.2.2;Network Design;94 11.2.3;Inputs and Output;95 11.3;Results;96 11.4;Conclusions;97 11.5;References;97 12;Information Enhancement Learning: Local Enhanced Information to Detect the Importance of Input Variables in Competitive Learning;98 12.1;Introduction;98 12.2;Theory and Computational Methods;99 12.2.1;Enhancement and Relaxation;99 12.2.2;Self-enhancing;100 12.2.3;Collective Enhancement;102 12.2.4;Local Enhancement;103 12.3;Results and Discussion;104 12.3.1;Artificial Data;104 12.4;Conclusion;107 12.5;References;108 13;Flash Flood Forecas
ting by Statistical Learning in the Absence of Rainfall Forecast: A Case Study;110 13.1;Introduction;110 13.2;Problem Statement;111 13.2.1;Flash Flood Forecasting;111 13.2.2;$Gardon dAnduze$ Flash Floods;112 13.2.3;Noise and Accuracy;113 13.3;Model Design;113 13.3.1;Definition of the Model;113 13.3.2;Model Selection;114 13.3.3;Training;115 13.4;Regularization;115 13.4.1;Weight Decay;115 13.4.2;Early Stopping;116 13.5;Results and Discussion;116 13.6;Conclusion;118 13.7;References;119 14;An Improved Algorithm for SVMs Classification of Imbalanced Data Sets;120 14.1;Introduction;120 14.2;Background;121 14.2.1;Support Vector Machines;121 14.2.2;Related Works;123 14.3;Boundary Elimination and Domination Algorithm;124 14.4;Experiments and Results;126 14.4.1;Experiment Methodology;126 14.4.2;Results;127 14.5;Conclusions;129 14.6;References;129 15;Visualization of MIMO Process Dynamics Using Local Dynamic Modelling with Self Organizing Maps;131 15.1;Introduction;131 15.2;Local Linear Modelling of Dynamics;132 15.2.1;Clustering Dynamics;132 15.2.2;Local Model Estimation;133 15.2.3;Retrieval;133 15.2.4;Visualization of Dynamics;134 15.3;Experimental Results;134 15.3.1;Industrial-Scale 4-Tank Model;134 15.3.2;Experiment Description;136 15.3.3;Model Training and Validation;136 15.3.4;Visualization of MIMO Dynamic Features;138 15.4;Conclusion;141 15.5;References;141 16;Data Visualisation and Exploration with Prior Knowledge;143 16.1;Introduction;143 16.2;Data Exploration;144 16.2.1;Standard GTM;144 16.2.2;Extension to Block GTM;146 16.2.3;Extension of GTM for Missing Data Using EM;147 16.2.4;Stabilising the EM Algorithm;147 16.2.5;Assessing Unsupervised Learning;147 16.3;Experiments on Artificial Data;148 16.4;Experiments on Geochemical Data;151 16.5;Conclusions;152 16.6;Future Work;153 16.7;References;153 17;Reducing Urban Concentration Using a Neural Network Model;155 17.1;Introduction;155 17.2;The Neural Network Model;156 17.3;Some Highlights of the Model;158 17.4;A Real Exam
ple;159 17.5;Conclusion;163 17.6;References;163 18;Dissimilarity-Based Classification of Multidimensional Signals by Conjoint Elastic Matching: Application to Phytoplanktonic Species Recognition;165 18.1;Introduction;165 18.2;Dissimilarity Measure for Multidimensional Signals by Conjoint Elastic Matching;166 18.2.1;Comparison of Two 1D Signals by the Classical Method $Dynamic Time Warping$;166 18.2.2;Neighborhood Restrictions of DTW Algorithm;168 18.2.3;Dissimilarity Measure of Positive Signals;169 18.2.4;Conjoint Elastic Matching of nD Signals;169 18.2.5;Matching Visualizations;169 18.3;Application to the Phytoplanktonic Species Identification;171 18.3.1;Data Presentation;171 18.3.2;Applied Classification Methods;173 18.3.3;Classification Results;174 18.4;Conclusion;175 18.5;References;176 19;Revealing the Structure of Childhood Abdominal Pain Data and Supporting Diagnostic Decision Making;177 19.1;Introduction;178 19.2;Random Forest Implementations;179 19.2.1;Experiment Setup;179 19.2.2;Experimental Results and Discussion;182 19.3;Genetic Algorithm Clustering and Genetic Feature Selection;184 19.3.1;Experiment Setup;185 19.3.2;Experimental Results and Discussion;186 19.4;Conclusions;187 19.5;References;187 20;Relating Halftone Dot Quality to Paper Surface Topography;190 20.1;Introduction;190 20.2;Data Acquisition;191 20.3;Analysis Methods;192 20.3.1;Self Organizing Maps;192 20.3.2;Clustering;193 20.3.3;Support Vector Machine Classification;193 20.4;Analysis of Print Quality;194 20.5;Analysis of Topography and Print;195 20.6;SVM Classification;197 20.6.1;C-SVC Classification;198 20.6.2;$.$-SVM Classification;198 20.6.3;Class Probabilities;199 20.7;Conclusion;199 20.8;References;200 21;Combining GRN Modeling and Demonstration-Based Programming for Robot Control;202 21.1;Introduction;202 21.2;Developing GRN Controllers;203 21.2.1;RNN-Based Regulatory Model;203 21.2.2;Learning Algorithm for Constructing GRN Controllers;205 21.2.3;Demonstration-Based Programming;206 21
.3;Experiments and Results;206 21.3.1;Modeling GRNs from Expression Data;207 21.3.2;Learning GRNs for Robot Control;208 21.4;Conclusions and Future Work;210 21.5;References;211 22;Discriminating Angry, Happy and Neutral Facial Expression: A Comparison of Computational Models;212 22.1;Introduction;212 22.2;Background;213 22.2.1;Gabor Filters;213 22.2.2;Curvilinear Component Analysis;215 22.2.3;Intrinsic Dimension;216 22.2.4;Classification Using Support Vector Machines;216 22.3;Experiments and Results;217 22.4;Conclusions;220 22.5;References;220 23;Modeling and Forecasting CAT and HDD Indices for Weather Derivative Pricing;222 23.1;Introduction;222 23.2;Modeling Temperature Process;224 23.3;Wavelet Neural Networks for Multivariate Process Modeling;225 23.4;Modeling and Forecasting CAT and HDD Indices;227 23.5;Temperature Derivative Pricing;230 23.6;Conclusions;232 23.7;References;233 24;Using the Support Vector Machine as a Classification Method for Software Defect Prediction with Static Code Metrics;235 24.1;Introduction;235 24.2;Background;236 24.2.1;Static Code Metrics;236 24.2.2;The Support Vector Machine;237 24.2.3;Data;237 24.3;Method;239 24.3.1;Data Pre-processing;239 24.3.2;Experimental Design;241 24.4;Assessing Performance;242 24.5;Results;242 24.6;Analysis;243 24.7;Conclusion;244 24.8;References;244 24.9;Appendix;246 25;Adaptive Electrical Signal Post-processing with Varying Representations in Optical Communication Systems;247 25.1;Introduction;247 25.2;Background to the Problem;248 25.3;Description of the Data;249 25.3.1;Representation of the Data;250 25.3.2;The Discrete Wavelet Transform;251 25.3.3;Independent Component Analysis;252 25.4;Method;252 25.4.1;Easy and Hard Cases;252 25.4.2;Visualisation Using PCA;253 25.4.3;Single Layer Neural Network;254 25.5;Performance Measures;254 25.6;Experiments;255 25.6.1;The First Experiment;255 25.6.2;The Second Experiment;255 25.6.3;The Third Experiment;256 25.7;Discussion;256 25.8;References;257 26;Using of Artifici
al Neural Networks (ANN) for Aircraft Motion Parameters Identification;258 26.1;Introduction;258 26.2;Longitudinal Forces during Takeoff Run;259 26.3;Problem Definition;260 26.4;Identification Procedure;261 26.5;Identification Results;261 26.6;Check Modeling;263 26.7;Estimation of Actual Aircraft Braking Characteristics under Different Runway Conditions;265 26.8;Further Solutions for ANN-Based Identification Tasks;267 26.9;References;268 27;Ellipse Support Vector Data Description;269 27.1;Introduction;269 27.2;Support Vector Data Description;271 27.3;The Proposed Method;272 27.3.1;$\Phi$ Functions Characteristics;276 27.4;Experiments;277 27.5;Conclusion;279 27.6;References;279 28;Enhanced Radial Basis Function Neural Network Design Using Parallel Evolutionary Algorithms;281 28.1;Introduction;281 28.2;State of the Art;282 28.3;Method Overview;283 28.3.1;Description of EvRBF;283 28.3.2;Description of Symbiotic_CHC_RBF;284 28.3.3;Description of SymbPar;284 28.4;Experiments and Results;286 28.5;Conclusions and Future Research;290 28.6;References;291 29;New Aspects of the Elastic Net Algorithm for Cluster Analysis;293 29.1;Introduction;293 29.2;Some New Aspects on the Elastic Net Algorithm;294 29.3;Application to Artificial Created Two-Dimensional Clusters;296 29.4;Conclusions;301 29.5;References;302 30;Neural Networks for Forecasting in a Multi-skill Call Centre;303 30.1;Introduction;303 30.2;Forecasting in Call Centres;304 30.3;Forecasting in CCs Using NNs;305 30.3.1;Background;305 30.3.2;Data Set;306 30.3.3;Variables;307 30.3.4;Metrics;307 30.3.5;Adaptations;308 30.3.6;Adaptive Learning Rate;308 30.4;Results;309 30.4.1;Analysis;309 30.4.2;Comparative;310 30.5;Conclusions;311 30.6;References;312 31;Relational Reinforcement Learning Applied to Appearance-Based Object Recognition;313 31.1;Introduction;313 31.2;Reinforcement Learning;314 31.3;Relational Reinforcement Learning;314 31.4;Appearance-Based Modeling;315 31.5;Application;318 31.6;Conclusion;323 31.7;Future Resea
rch;323 31.8;References;324 32;Sensitivity Analysis of Forest Fire Risk Factors and Development of a Corresponding Fuzzy Inference System: The Case of Greece;325 32.1;Introduction;325 32.2;Theoretical Framework;326 32.3;Pearson Correlation Coefficients;329 32.4;Sensitivity Analysis;330 32.4.1;Analysis Related to the Case of Forest Fire Incidents;330 32.4.2;Analysis Related to the Case of the Burned Area;332 32.5;Compatibility of the Systems Output to the Actual Case;334 32.6;Conclusions and Discussion;335 32.7;References;335 33;Nonmonotone Learning of Recurrent Neural Networks in Symbolic Sequence Processing Applications;337 33.1;Introduction;337 33.2;Nonmonotone Training Algorithms;339 33.3;Experiments and Results;341 33.4;Conclusions;346 33.5;References;346 34;Indirect Adaptive Control Using Hopfield-Based Dynamic Neural Network for SISO Nonlinear Systems;348 34.1;Introduction;348 34.2;Hopfield-Based Dynamic Neural Model;349 34.2.1;Descriptions of the DNN Model;349 34.2.2;Hopfied-Based DNN Approximator;350 34.3;Problem Formulation;351 34.4;Design of IACHDNN;352 34.5;Simulation Results;359 34.6;Conclusions;360 34.7;References;361 35;A Neural Network Computational Model of Visual Selective Attention;362 35.1;Introduction;362 35.2;Proposed Computational Model of Visual Selective Attention;364 35.3;Coincidence Detector Neurons and the Correlation Control Module;365 35.4;Simulations and Evaluation of the Model;367 35.4.1;Attentional Blink Explanation Theory;367 35.5;Discussion;369 35.6;References;369 36;Simulation of Large Spiking Neural Networks on Distributed Architectures, The DAMNED Simulator;371 36.1;MIMD-DM Architectures;372 36.2;Architecture of DAMNED;372 36.3;Delayed Queues of Events;374 36.4;Conservative and Distributed Virtual Clock Handling;375 36.5;Configuration of DAMNED and Definition of a SNN;378 36.6;Results;378 36.7;Conclusion and Future Work;380 36.8;References;381 37;A Neural Network Model for the Critical Frequency of the F2 Ionospheric Layer over C
yprus;383 37.1;Introduction;383 37.2;Characteristics of the F2 Layer Critical Frequency;384 37.3;Model Parameters;386 37.4;Experiments and Results;387 37.5;Conclusions and Future Work;389 37.6;References;389 38;Dictionary-Based ClassificationModels. Applications for Multichannel Neural Activity Analysis;390 38.1;Introduction;390 38.2;Material and Methods;392 38.2.1;Animal Training and Behavioral Tasks;392 38.2.2;Chronic Animal Preparation and Neural Ensemble Recording;392 38.3;Data Analysis;392 38.3.1;Preprocessing;392 38.3.2;Manual Scatterplot Classification;393 38.3.3;Quantification and Classification of Spike Waveforms;393 38.4;Definition of SDC;397 38.5;Results;398 38.6;Conclusion;399 38.7;References;399 39;Pareto-Based Multi-output Metamodeling with Active Learning;401 39.1;Introduction;401 39.2;Global Surrogate Modeling;402 39.3;Multi-objective Modeling;402 39.4;Related Work;403 39.5;Problems;404 39.5.1;Analytic Function;404 39.5.2;Low Noise Amplifier (LNA);405 39.6;Experimental Setup;405 39.6.1;SUMO-Toolbox;405 39.6.2;Analytic Function (AF);406 39.6.3;LNA;406 39.7;Results;407 39.7.1;Analytic Function: Use Case 1;407 39.7.2;Analytic Function: Use Case 2;408 39.7.3;LNA: Use Case 1;409 39.7.4;LNA: Use Case 2;410 39.8;Conclusion and Future Work;411 39.9;References;411 40;Isolating Stock Prices Variation with Neural Networks;413 40.1;Introduction;413 40.2;Literature Review;414 40.3;Experiments and Results;416 40.3.1;Experimental Methodology;416 40.3.2;Results;417 40.4;Conclusions;419 40.5;References;420 41;Evolutionary Ranking on Multiple Word Correction Algorithms Using Neural Network Approach;421 41.1;Introduction;421 41.2;Typing Correction Functions;422 41.3;Word List Neural Network Ranking and Definitions;423 41.4;Word List Neural Network Ranking Modelling;425 41.5;Conclusion;430 41.6;References;430 42;Application of Neural Fuzzy Controller for Streaming Video over IEEE 802.15.1;431 42.1;Introduction;431 42.2;Methodology;432 42.3;Computer Simulation Results;43
5 42.4;Conclusion;440 42.5;References;440 43;Tracking of the Plasma States in a Nuclear Fusion Device Using SOMs;442 43.1;Introduction;442 43.2;Data Visualization with SOM;443 43.2.1;Quality Indexes;444 43.3;The Data Base Composition;445 43.4;Results;446 43.5;Conclusions;449 43.6;References;449 44;An Application of the Occam Factor to Model Order Determination;450 44.1;Introduction;450 44.2;The Bayesian Approach;450 44.3;Formulating the Error Term E$_{D}$(w) for the Simple Polynomial Model;451 44.4;Deriving a Second Order Approximation for the Log Posterior;452 44.5;Calculation of the Evidence P (D|HK);452 44.6;Conclusions and Future Work;455 44.7;References;455 45;Use of Data Mining Techniques for Improved Detection of Breast Cancer with Biofield Diagnostic System;456 45.1;Introduction;456 45.2;Description of the Datasets;458 45.2.1;Dataset 1;458 45.2.2;Dataset 2;458 45.3;Proposed Data Mining Framework;459 45.4;Results and Discussion;461 45.4.1;TTSH Dataset Classification Results;461 45.4.2;US Dataset Classification Results;462 45.4.3;Comparison of Results and Discussion;462 45.5;Conclusion;463 45.6;References;464 46;Clustering of Entropy Topography in Epileptic Electroencephalography;465 46.1;Introduction;465 46.2;The Electroencephalography and the Neuronal Sources;467 46.3;Methodology;468 46.3.1;Entropy to Measure Order in the Brain;468 46.3.2;Entropy Topography Clustering;469 46.4;Results;470 46.4.1;Data Description;470 46.4.2;Movie Visual Review;470 46.4.3;Electrodes Clustering;470 46.5;Conclusions;472 46.6;References;473 47;Riverflow Prediction with Artificial Neural Networks;475 47.1;Introduction;475 47.2;Daily Riverflow Prediction;476 47.2.1;Mekong River at Pakse Gauging Station, Lao;476 47.2.2;Principal Component Analysis;477 47.2.3;Chao Phraya River at Nakhon Sawan Gauging Station, Thailand;478 47.3;Daily Stage Prediction;480 47.3.1;Surma River at Sylhet Gauging Station, Bangladesh;480 47.4;Performance Criteria;481 47.5;Concluding Remarks;482 47.6;Referenc
es;482 48;Applying Snap-Drift Neural Network to Trajectory Data to Identify Road Types: Assessing the Effect of Trajectory Variability;484 48.1;Introduction;484 48.2;Past Work on Trajectory Analysis;485 48.3;Snap-Drift Neural Networks;486 48.3.1;The Snap-Drift Algorithm;488 48.4;Data Collection;489 48.5;Road Design Parameter Derivation from GPS Trajectory Data;490 48.6;Trajectory Data Variability Analysis;490 48.6.1;Trajectory Variability Reduction;492 48.7;Data Types;492 48.8;Results;492 48.9;Conclusion;494 48.10;References;495 49;Reputation Prediction in Mobile Ad Hoc Networks Using RBF Neural Networks;497 49.1;Introduction;497 49.2;Related Work;499 49.3;Modeling the Network;500 49.4;Simulation Details;501 49.5;Simulation Results;503 49.6;Practical Considerations;505 49.7;Conclusions and Future Work;505 49.8;References;506 50;Author Index;507


EAN: 9783642039690
Untertitel: 11th International Conference, EANN 2009, London, UK, August 27-29, 2009, Proceedings. Sprache: Englisch. Dateigröße in MByte: 23.
Verlag: Springer Berlin Heidelberg
Erscheinungsdatum: August 2009
Format: pdf eBook
Kopierschutz: Adobe DRM
Es gibt zu diesem Artikel noch keine Bewertungen.Kundenbewertung schreiben