Information Technology Reference
In-Depth Information
between the initial
C
at the start of an independent run and the final
C
at the end of the
independent run, it can be defined as
. The
△
C
also indicates the
increasing on the number of false negatives of the strategies.
To prove that the detector sets being updated by the new strategies are less affected
by the growing self set, and have a better coverage on the non-self space, comparisons
on
△
C
and changes of
N
R
after detector sets have been updated by the three strategies
are given in the following two subsections. In subsection 4.1, the three strategies are
compared when initial
N
R
is fixed, and it can be found out that the
△
C
and the change
on
N
R
of the new strategies are smaller than that of strategy 0. In subsection 4.2, the
comparison in subsection 4.1 are repeated when the self set is unavailable during the
updating process of detector set, and similar results are found.
Δ
C
=
C
−
C
initial
final
4.1 Comparisons When Initial
N
R
Is Fixed
In this subsection, the experiment is carried out to make the comparisons on
C
when
three strategies have the same initial number of detectors (
N
R
). The initial
N
R
is set to
6000 in this experiment.
△
Table 1.
Comparison on average
△
C
over 10 independent runs when initial
N
R
is fixed to 6000.
Standard deviations are also listed in this table.
a
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
Ave 3696.9
7359.7 10799.0 14481.0
17301.0 20680.0
23152.0 25968.0 28002.0
0
Std
315.38
365.71 230.93
359.13
426.68
358.01
819.97
990.19
517.75
Ave 2572.2
5059.1 7142.5 9176.5 10788.0 12395.0
13779.0 15120.0 16279.0
I
Std
173.71
228.14 133.40
211.05
204.12
225.29
406.82
401.01
338.85
Ave 1002.2
1725.5 2133.8
2850.7
3382.9.0
3930.0
4564.2.0
5477.0
6152.0
II
Std
101.63
198.10 259.85
314.03
282.94
372.59
247.63
242.43
173.06
Table 1 shows the comparison between the three strategies on average
△
C
, the
standard deviations are also listed. It can be observed that the values of
C
in the
results of strategy I and II are lower
than that of strategy 0. It means that the increasing
on false negative ratio of these two novel strategies is lower than that of strategy 0, i.e.
the non-self space coverage of the detector sets being updated by these two new
strategies are less affected by the growing of self set. And it can be also seen that
strategy II is much better than strategy I from the
△
C
's point of view.
Table 2 lists the average values of final
N
R
of three strategies after 10 independent
runs against values of
a
, the average numbers of deleted detectors and regenerated
(rebirth) detectors are also given, and standard deviations are listed as well. It can be
found out that more existent detectors are reserved by strategy I and II during the
detector set updating process, i.e. the detector set being updated by the two new
strategies are less affected by the growing of self set. And strategy I is a little better than
strategy II from the
N
R
's point of view.
△
