Implementation techniques / edited by Cornelius T. Leondes.

Format:

Book

Contributor:

Leondes, Cornelius T.

Series:

Neural Network Systems Techniques and Applications

Neural network systems, techniques, and applications ; 3

Language:

English

Subjects (All):

Neural networks (Computer science).

Natural computation.

Physical Description:

1 online resource (421 p.)

Place of Publication:

San Diego : Academic Press, c1998.

Language Note:

English

Summary:

This volume covers practical and effective implementation techniques, including recurrent methods, Boltzmann machines, constructive learning with methods for the reduction of complexity in neural network systems, modular systems, associative memory, neural network design based on the concept of the Inductive Logic Unit, and a comprehensive treatment of implementations in the area of data classification. Numerous examples enhance the text. Practitioners, researchers,and students in engineering and computer science will find Implementation Techniques a comprehensive and powerful reference

Contents:

Cover; Title Page; Copyright Page; Contents; Contributors; Preface; Chapter 1. Recurrent Neural Networks: Identification and Other System Theoretic Properties; I. Introduction; II. Recurrent Neural Networks; III. Mixed Networks; IV. Some Open Problems; References; Chapter 2. Boltzmann Machines: Statistical Associations and Algorithms for Training; I. Introduction; II. Relationship with Markov Chain Monte Carlo Methods; III. Deterministic Origins of Boltzmann Machines; IV. Hidden Units; V. Training a Boltzmann Machine; VI. An Example with No Hidden Unit; VII. Examples with Hidden Units

VIII. Variations on the Basic Boltzmann MachineIX. The Future Prospects for Boltzmann Machines; References; Chapter 3. Constructive Learning Techniques for Designing Neural Network Systems; I. Introduction; II. Classification; III. Regression Problems; IV. Constructing Modular Architectures; V. Reducing Network Complexity; VI. Conclusion; VII. Appendix: Algorithms for Single-Node Learning; References; Chapter 4. Modular Neural Networks; I. Introduction; II. Why Modular Networks?; III. Modular Network Architectures; IV. Input Decomposition; V. Output Decomposition

VI. Hierarchical DecompositionVII. Combining Outputs of Expert Modules; VIII, Adaptive Modular Networks; IX. Conclusions; References; Chapter 5. Associative Memories; I. Introduction; II. Point Attractor Associative Memories; III. Continuous PAAM: Competitive Associative Memories; IV. Discrete PAAM: Asymmetric Hopfield-Type Networks; V. Summary and Concluding Remarks References; Chapter 6. A Logical Basis for Neural Network Design; I. Motivation; II. Overview; III. Logic, Probability, and Bearing; IV. Principle of Maximized Bearing and ILU Architecture; V. Optimized Transmission

VI. Optimized TransductionVII. ILU Computational Structure; VIII. ILU Testing; IX. Summary; Appendix: Significant Marginal and Conditional ILU Distributions; References; Chapter 7. Neural Networks Applied to Data Analysis; I. Introduction; II. Data Complexity; III. Data Separability; IV. Classifier Selection; V. Classifier Nonlinearity; VI. Classifier Stability; VII. Conclusions and Discussion; References; Chapter 8. Multimode Single-Neuron Arithmetics; I. Introduction; II. Defining Neuronal Arithmetics; III. Phase Space of Neuronal Arithmetics; IV. Multimode Neuronal Arithmetic Unit

V. Toward a Computing NeuronVI. Summary; References; Index

Notes:

Description based upon print version of record.

Includes bibliographical references and indexes.

ISBN:

1-281-05427-5

9786611054274

0-08-055182-3

OCLC:

476103120