Modular Learning in Neural Networks
A Modularized Approach to Neural Network Classification
Published on 9. October 1992
Book
Hardback
256 pages
978-0-471-57154-4 (ISBN)
Description
The author attempts to depart from conventional thinking by viewing neural networks as structured systems instead of monoliths. He presents various ways in which the understanding of neural networks can be broken down into relatively independent parts, providing intermediary solution stages.
More details
Series
Language
English
Place of publication
New York
United States
Publishing group
John Wiley and Sons Ltd
Target group
Professional and scholarly
Illustrations
Ill.
Dimensions
Height: 242 mm
Width: 161 mm
Weight
567 gr
ISBN-13
978-0-471-57154-4 (9780471571544)
Copyright in bibliographic data is held by Nielsen Book Services Limited or its licensors: all rights reserved.
Schweitzer Classification
Content
NEURAL NETWORKS; Introduction to Neural Networks; Structure of the Brain; MODULAR LEARNING; Decomposition of Learning Into Unsupervised and Supervised Learning; Supporting Supervised Learning by Feature Extraction; Supporting Supervised Learning by Quantization; Finding Optimal Features for a Given Task; Decomposition of the Represented Mapping; Decomposing the Network to Minimize Interactions; Modularizing the Application Task; Decomposing Network Construction Into Knowledge-Based and Learning Parts; Conclusions; References; Index.