View Table of ContentsCreate a BookmarkAdd Book to My BookshelfPurchase This Book Online
Skip to Book Content
Book cover image

Index

Neural Smithing: Supervised Learning in Feedforward Artificial Neural Networks
Russell D. Reed and Robert J. Marks II
Copyright © 1999 Massachusetts Institute of Technology

Previous Section Next Section

Index

M

Marchand's algorithm, 209-212
Master units, 206-209
Mean squared error (MSE) function, 9, 50, 155
dynamic node creation and, 200
error surface and, 117, 123, 129
generalization and, 239, 243, 253, 259, 261, 272, 276
initialization and, 110
learning rate and, 71, 73, 77, 139
pattern weighting and, 151
Widrow-Hoff learning rule and, 29-30
Meiosis networks, 212-213
Meta-optimization, 157
Minima. See Local minima
Mirror symmetry problem, 211
Model mismatch, 248
Modularity, 255
Momentum, 62-63
algorithm variations and, 137, 140, 142, 144
classical optimization and, 165
frequency response and, 93
gradient correlation and, 150-151
impulse response and, 91-92, 94
inertia and, 89
learning rate and, 71, 72-77, 80, 87-90, 92
small-signal analysis and, 90-95
step response and, 92, 94-95
training time and, 74-77, 85-95
values for, 71
Multilayer networks, importance of, 19
Multilayer perceptron (MLP)
as cascade of single-layer perceptrons, 31
compared to real nervous systems, 3-4
component layers of, 31
definition of, 29
introduction to, 2-6
representational capabilities of, 31-47
universal approximation by, 35-38
Mutation, 186, 187, 188-189

Top of current section
Previous Section Next Section
Books24x7.com, Inc. © 1999-2001  –  Feedback