Handbook of Neural Computing Applications is a collection of articles that deals with neural networks. Some papers review the biology of neural networks, their type and function (structure, dynamics, and learning) and compare a back-propagating perceptron with a Boltzmann machine, or a Hopfield network with a Brain-State-in-a-Box network. Other papers deal with specific neural network types, and also on selecting, configuring, and implementing neural networks. Other papers address specific applications including neurocontrol for the benefit of control engineers and for neural networks researchers. Other applications involve signal processing, spatio-temporal pattern recognition, medical diagnoses, fault diagnoses, robotics, business, data communications, data compression, and adaptive man-machine systems. One paper describes data compression and dimensionality reduction methods that have characteristics, such as high compression ratios to facilitate data storage, strong discrimination of novel data from baseline, rapid operation for software and hardware, as well as the ability to recognized loss of data during compression or reconstruction. The collection can prove helpful for programmers, computer engineers, computer technicians, and computer instructors dealing with many aspects of computers related to programming, hardware interface, networking, engineering or design.



Inhalt

Acknowlegments
Preface

1 Introduction to Neural Networks

1.0 Overview

1.1 Practical Applications

1.2 The Advantages of Neural Networks

1.3 A Definition of Neural Networks

1.4 Summary

References

2 History and Development of Neural Networks

2.0 Overview

2.1 Early Foundations

2.2 Promising and Emerging Technology

2.3 Disenchantment

2.4 Innovation

2.5 Re-Emergence

2.6 Current Status

2.7 Summary

References

3 The Neurological Basis for Neural Computations

3.0 Neuroscience As A Model

3.1 The Single Neuron

3.2 Early Research

3.3 Structural Organization of Biological Neural Systems

3.4 Structurally Linked Dynamics of Biological Neural Systems

3.5 Emergent Properties Arise from the Dynamics of Biological Neural Systems

3.6 Learning in Biological Neural Systems

3.7 Functional Results of Neural Architecture

3.8 Computer Simulations Based on the Brain

References

4 Neural Network Structures: Form Follows Function

4.0 Overview

4.1 Levels of Structural Description

4.2 Neural Micro-Structures

4.3 Neural Meso-Structures

4.4 The Macro-Structure

4.5 Summary

5 Dynamics of Neural Network Operations

5.0 Overview

5.1 Typical Network Dynamics

5.2 Energy Surfaces and Stability Criterion

5.3 Network Structures and Dynamics

References

6 Learning Background for Neural Networks

6.0 Overview

6.1 Intelligence: An Operational Definition

6.2 Learning and Conditioning

6.3 Learned Performance

6.4 Motivation

6.5 Summary

References

7 Multilayer Feedforward Neural Networks I: Delta Rule Learning

7.0 Overview

7.1 Introduction

7.2 The Perceptron Network

7.3 Adaline and Madaline Neural Networks

7.4 The Back-Propagation Network

References

8 Multilayer Feedforward Neural Networks II: Optimizing Learning Methods

8.0 Overview

8.1 The Boltzmann Machine

8.2 The Cauchy Machine: A Refinement of the Boltzmann Machine

8.3 Summary

References

9 Laterally-Connected, Autoassociative Networks

9.0 Overview

9.1 Introduction to Association Networks

9.2 Auto Associative Networks

9.3 The Hopfield/Tank Network

9.4 The Brain-State-In-A-Box Network

9.5 Kanerva's Sparse Distributed Memory Network

9.6 Summary

References

10 Vector-Matching Networks

10.0 Overview

10.1 Introduction

10.2 The Kohonen Learning Vector Quantization Network

10.3 The Self-Organizing Topology-Preserving Map

10.4 Summary

References

11 Feedforward/Feedback (Resonating) Heteroassociative Networks

11.0 Chapter Overview

11.1 Introduction

11.2 The Carpenter/Grossberg Adaptive Resonance Theory Network

11.3 Bidirectional Associative Memories and Related Networks

11.4 Summary

References

12 Multilayer Cooperative/Competitive Networks

12.0 Overview

12.1 Introduction

12.2 Competitive Learning Networks

12.3 Masking Fields

12.4 The Boundary Contour System

12.5 Hierarchical Scene Structures

12.6 The Neocognitron

12.7 Summary

References

13 Hybrid and Complex Networks

13.0 Overview

13.1 Introduction

13.2 Hybrid Networks: The Hamming Network and the Counter-Propagation Network

13.3 Neural Networks Operating in Parallel

13.4 Hierarchies of Similar Networks

13.5 Systems of Different Types of Neural Networks

13.6 Systems of Networks are Useful for Adaptive Control

13.7 Summary

References

14 Choosing A Network: Matching the Architecture to the Application

14.0 Chapter Overvie

14.1 When to use A Neural Network

14.2 What Type of Network?

14.3 Debugging, Testing, and Verifying Neural Network Codes

14.4 Implementing Neural Networks

References

15 Configuring and Optimizing the Back-Propagation Network

15.0 Overview

15.1 Issues in Optimizing and Generalizing Feedforward Networks

15.2 Micro-Structural Considerations

15.3 Meso-Structural Considerations

15.4 Optimizing Network Dynamics

15.5 Learning Rule Modifications

15.6 Modifications to Network Training Schedules and Datasets

References

16 Electronic Hardware Implementations

16.0 Overview

16.1 Analog Implementations

16.2 Digital Neural Network Chips

16.3 Hybrid Neural Network Chips

16.4 Method for Comparing Neural Network Chips

16.5 Summary

Further Reading in Neural Network Hardware Implementation

17 Optical Neuro-Computing

17.0 Overview

17.1 Historical Introduction of Optical Neurocomputing

17.2 Review of Learning Algebras and Architectures

17.3 Associative Memory vs. Wiener Filter and Self-Organization-Map vs. Kalman Filters

17.4 Optical Implementations of Neural Networks

17.5 Comparison Between Electronic and Optic Implementations of Neural Networks

17.6 Hybrid Neurocomputing

17.7 Application to Pattern Recognition and Image Processing

17.8 The Superconducting Mechanism

17.9 The Super-Triode

17.10 The Super-Triode Neurocomputer

17.11 Wave-Front Imaging Telescope with a Focal Plane Array of Super-Triodes

17.12 Space-Borne In-Situ Smart Sensing with Neurocomputing

17.13 Conclusion

Bibliography

18 Neural Networks for Spatio-Temporal Pattern Recognition

18.0 Overview

18.1 Creating Spatial Analogues of Temporal Patterns

18.2 Neural Networks with Time Delays

18.3 Storing and Generating Temporal Patterns Via Recurrent Connections

18.4 Using Neurons with Time-Varying Activations and Summing Information Over Time Intervals

18.5 Neural Nets which have Short-Term and Long-Term Memories

18.6 Frequency Coding in Neural Networks

18.7 Networks with Combinations of Different Tempor…

Titel
Handbook of Neural Computing Applications
EAN
9781483264844
Format
E-Book (pdf)
Veröffentlichung
10.05.2014
Digitaler Kopierschutz
Wasserzeichen
Dateigrösse
34.17 MB
Anzahl Seiten
470