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

L

Laplace transforms, 91, 92, 94
Layered architecture, introduction to, 2-6
Layers
components of, 31
counting of, 31
delta attenuation in, 72, 84-85
one hidden, sufficiency of, 33-41
representational capabilities and, 32-47
two hidden, sufficiency of, 32-33, 38-39, 109
Learning rate, 57, 62
additive versus multiplicative changes in, 140-141
algorithm variations and, 135-151
batch, 77-79
classical optimization and, 164
delta attenuation and, 84-85
effective, 77, 87, 88, 90
example for, 73-74
gain scaling and, 132-133
momentum and, 71, 72-77, 80, 87-90, 92
MSE function and, 71, 73
on-line, 80
randomness and, 80
scaling of, 85
SSE function and, 71
training time and, 71-85
values for, 71
Learning rate selection
adaptive methods and, 95
factors affecting, 71-72
momentum and, 89
from trace(H), 81-82
Learning rules, 23-30. See also Algorithms
Least mean squares (LMS) algorithm, 29, 298. See also Widrow-Hoff learning rule
error function and, 123, 124
momentum and, 90
Leave-one-out method, 257
Levenberg-Marquardt method, 128, 144, 172-173, 182
Linear discriminant analysis, 306-309
Linear output networks, 283-285
Linear regression, 23, 29, 293-298
Linear separability, 18-23, 28, 124
Linear threshold units (LTUs), 15, 23, 113
constructive methods and, 204-209
pruning and, 232
Line search, 158-159
Loading problem, 197
Local bottlenecks, 232-233
Local minima, 115-117, 121-126, 157
Logarithmic error function, 9, 50. See also Cross-entropy error function
Logistic regression, 23
LVQ method, 107

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