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

P

PAC (probably approximately correct) learning, 261-263
Paralysis, from sigmoid saturation, 68, 70
Parametric classifier, 152-153
Parametric models, 248
Pattern-mode learning, 62
Pattern moves, 160
Pattern weighting heuristics, 151
Penalty terms, 10-11, 220-221, 226-231, 237
Perceptron learning algorithm, 23-24
constructive methods and, 202, 205, 210-212
convergence proof for, 25-27
Perceptrons
definitions for, 1-2, 28-29
multilayer (see Multilayer perceptron)
Rosenblatt's original, 28-29
Perturbation, 56
Pinnacle function, 38
Pocket algorithm, 28, 205
Polak-Ribiere form, 167
Powell's conjugate direction method, 161-163
Prediction systems, and generalization, 257-264
Principal components analysis (PCA), 299-302
autoencoder networks and, 303-305
discriminant analysis and, 306-310
for initialization, 106
for pruning, 234-235
unsupervised learning and, 12
Processing units, 1
Projection pursuit regression, 217
Projections, hyperplane, 15
Prototypes, 108, 207-208
Pruning methods
bottlenecks and, 232-234
constructive methods and, 198, 201
example for, 219
gain competition, 232-233
generalization and, 219, 237, 268
genetic algorithm, 193, 235
interactive, 232
optimal brain damage, 131, 223, 225, 235
optimal brain surgeon, 223-225, 235
penalty terms, 220-221, 226-231, 237
principal components, 234-235
sensitivity, 220, 221-225, 237
skeletonization, 221-222
weight decay, 230-231
weight elimination, 228-229

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