One-hidden-layer networks,
33-
41,
124
algorithm variations and,
147
eigenvalue estimation and,
82-
84
On-line learning (cont.)
pseudocode example for,
65-
66
random initialization and,
105
training time and,
68,
80
Outer-product approximation,
130-
131
derivative calculation for,
53-
54
random initialization and,
102,
104
Overtraining
constructive methods and,
198
learning dynamics and,
249-
251