Information Technology Reference
In-Depth Information
2
v
⎡
2
v
⎤
(
)
(
)
∑
∑
ρ
=
μ
X
,
X
ρ
=
μ
M
,
M
;
⎢
⎣
⎥
⎦
j
R
i
j
j
R
i
j
1
1
i
=
1
i
=
1
2
v
⎡
2
v
⎤
(
)
(
)
∑
∑
ρ
=
μ
X
,
X
ρ
=
μ
M
,
M
.
(4.15)
⎢
⎣
⎥
⎦
j
R
i
j
j
R
i
j
2
2
i
=
1
i
=
1
(
)
and,
considering that an element corresponding to this index belongs to the first cluster
of potency
v
, we obtain a conditional ordered series of elements
ρ
ρ
,
j
=
1
2
v
j
=
1
2
v
Let us select the maximum index from indexes
,
j
j
k
Ξ
(or elements
k
Θ
)
X
=
X
=
...
=
X
>
X
=
X
=
...
=
X
>
X
>
...
>
X
.
(
)
( )
1
2
v
v
+
1
v
+
2
2
v
2
v
+
1
k
Individual clusters are ranged similarly to ranging of individual clusters in p. 3. In
this case it is easy to demonstrate that
2
k
−
v
+
1
2
k
−
3
v
+
1
ω
=
,
i
=
1
v
;
ω
=
,
i
=
v
+
1
2
v
;
(
)
i
(
)
k
k
+
1
i
k
k
+
1
(
)
2
k
−
i
+
1
ω
=
,
i
=
2
v
+
1
k
;
(4.16)
()
(
)
i
k
k
+
1
i
=
2
v
+
1
k
b) A set of elements with indexes
is fractionalized into clusters
of different potency. The potency of each cluster is less than
v
, but it is
more than '1' or equal to '1'. In this case the most powerful cluster is
selected, and the procedure from p. 4 is
carried
out to determine weight
coefficients of elements with indexes
. Weight coefficients of
elements of clusters of potency
v
remain unchanged
i
=
2
v
+
1
k
2
k
−
v
+
1
2
k
−
3
v
+
1
ω
=
,
i
=
1
v
;
ω
=
,
i
=
v
+
1
2
v
;
(4.17)
(
)
(
)
i
i
k
k
+
1
k
k
+
1
c) The set of elements with indexes
is fractionalized into clusters
with potencies less than
v
each, but more or equal to '1'y, and among
these clusters there are several large ones identical in their potencies. Let
us consider t
hat there are t
wo clusters of potency
i
=
2
v
+
1
k
d
>
1
with elements
(
)
X
j
M
j
=
2
v
+
1
2
v
+
2
d
ρ
ρ
,
. For these elements indexes
or
are
j
j
j
defined.
v
⎡
v
⎤
(
)
(
)
∑
∑
ρ
=
μ
X
,
X
ρ
=
μ
M
,
M
;
⎢
⎣
⎥
⎦
j
R
i
j
j
R
i
j
1
1
i
=
1
i
=
1
v
⎛
v
⎞
(
)
(
)
∑
∑
ρ
=
μ
X
,
X
⎜
⎝
ρ
=
μ
M
,
M
⎟
⎠
.
(4.18)
j
R
i
j
j
R
i
j
2
2
i
=
1
i
=
1
