SAB (self-adapting back propagation),
142
Scaled conjugate gradient descent,
168
Search then converge method,
147-
148
for single-layer networks,
15
Sigmoid saturation,
68,
69,
70
algorithm variations and,
141
ill-conditioning and,
128
random initialization and,
97,
102
Single-hidden-layer networks,
33-
41,
124
algorithm selection and,
157
constructive methods and,
197,
199
in perceptron learning algorithm,
24
for single-layer networks,
15,
24,
29
Structure
constructive methods and,
197-
217
Subpopulations, and genetic algorithm,
192-
193
Sum of squared errors (SSE) function,
9,
50,
52,
54
learning rate and,
71,
77
Symmetries, weight-space,
118-
119
System energy, and simulated annealing,
177