Information Technology Reference
In-Depth Information
Table 6.4
(
continued
)
8
(0,45;0,55;0,3;0,3)
(0;0,125;0;0,25)
(0,375;0,55;0,25;0,3)
(0;0,075;0;0,15)
9
(0,45;0,55;0,3;0,3)
(0,775;1;0,35;0)
(0;0,125;0;0,25)
(0,225;0,675;0,15;0,25)
10
(0,45;0,55;0,3;0,3)
(0,775;1;0,35;0)
(0;0,125;0;0,25)
(0,225;0,675;0,15;0,25)
11
(0;0,15;0;0,3)
(0;0,125;0;0,25)
(0,375;0,55;0,25;0,3)
(0;0,075;0;0,15)
12
(0,45;0,55;0,3;0,3)
(0,375;0,425;0,25;0,35)
(0;0,125;0;0,25)
(0;0,075;0;0,15)
It is assumed that all characteristics have three linguistic values: "low", "mean",
"high". The choice of input characteristics from a certain set offered is an
individual separate problem [211] aimed at detecting those which have essential
impact on the output characteristic.
Twelve selected software products have been estimated by experts, and the
obtained data are formalized by the method described in §2.2. Results are
summarized in Table 6.4.
By the method described in § 6.5, nonlinear hybrid fuzzy least-squares
regression model is constructed:
(
~
~
~
)
(
)
Y
0
483
;
X
2
1
0
061
;
X
2
3
=
−
+
+
~
~
~
(
)
(
)
+
−
0
121
;
X
X
+
1
022
;
022
;
X
+
2
3
1
~
(
)
(
)
+
0
283
;
283
;
X
+
0
017
;
.
By the method described in §6.4, linear hybrid fuzzy least-squares regression
model is constructed:
2
~
~
(
)
(
)
Y
=
0
026
;
+
0
067
;
X
+
1
~
~
(
)
(
)
X
++
On carrying out of the comparative analysis of real output data and the model
output data obtained within the limits of two constructed models, the nonlinear
regression model is selected to enable further building process.
Based on the data of Table 6.4, data
0
619
;
585
;
507
0
234
;
X
.
2
3
X
,
=
i
of the software products
obtaining the higher marks of success (tab. 6.5), is determined.
1
Table 6.5
Data of the software products obtaining the higher marks of success
X
X
X
1
2
3
(0,45;0,55;0,3;0,3)
(0;0,125;0;0,25)
(0,85;1;0,3;0)
(0,85;1;0,3;0)
(0;0,125;0;0,25)
(0,85;1;0,3;0)
(0,85;1;0,3;0)
(0,375;0,425;0,25;0,35)
(0,375;0,55;0,25;0,3)
For the data of Table 6.5 the weighed segments (tab. 6.6) is obtained.
