Information Technology Reference
In-Depth Information
14.00
1.00
12.00
.80
10.00
.60
8.00
6.00
.40
4.00
.20
2.00
Attributes
Adequately predicted by heuristic model
Remaining
.00
.00
.00
.50
1.00
1.50
2.00
2.50
3.00
.00
2.00
4.00
6.00
8.00
10.00
Estimated noise in attribute's ratings (sigma k)
Attribute variance
Fig. 3.5
(left) Scatter plot of the accounted variance by the three diverse models of the heuris-
tic technique versus the variance in the original attribute ratings, and (right) scatter plot of the
range versus the estimated noise in predicted attribute ratings.
1.00
1.00
.80
.80
.60
.60
.40
.40
.20
.20
.00
.00
.00
.20
.40
.60
.80
1.00
.00
.20
.40
.60
.80
1.00
Rsq - Algorithmic Technique
Rsq - Algorithmic Technique
14.00
25.00
12.00
20.00
10.00
15.00
8.00
6.00
10.00
4.00
5.00
2.00
.00
.00
.00
5.00
10.00
15.00
20.00
25.00
.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
Rk - Algorithmic Technique
Rk - Algorithmic Technique
Fig. 3.6
R
2
(top) and R
k
(bottom) values for the 38 attributes that were adequately modeled
by the three models of the heuristic technique proposed in this chapter (left) and the remain-
ing attributes (right), for the heuristic technique and the algorithmic technique proposed by
Martens (2009a).
Search WWH ::
Custom Search