Graphics Reference
In-Depth Information
Figure
.
.
Adaptive weights smoothing for a simple univariate regression problem: data (upper let),
regression function, and estimates θ
(
k
)
for k
) (upper
right). he lower part displays the contributions from the location penalty (top row) and the stochastic
penalty (middle row)totheweights(bottom row) w
ij
in iteration steps k
=
(h
=
), k
=
(h
=
) and k
=
(h
=
,
,
and
(columns).
=
A linear grayscale is provided at the bottom