Information Technology Reference
In-Depth Information
Table 2.
The average values of the final
N
R
over 10 independent runs against values of
a
when
initial
N
R
is fixed to 6000. The average numbers of deleted and regenerated (rebirth) detectors are
also given in this table. Standard deviations are listed as well.
a
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
Ave 1254.6 2240.7 3110.2
3769.0
4452.2
5100.9
5809.9
6560.7 7484.0
Deleted
Std
92.292 76.472 69.624
56.115
45.013
99.669
41.439
75.025 41.355
Ave 56.1 156.4 335.2 512.4 796.2 1148.5
1620.2
2217.1 3054.9
0
Rebirth
Std
7.125 10.146 11.114
20.764
41.787
55.838
46.913
50.516 58.872
Ave 4801.5 3915.7 3225.0
2743.4
2344.0
2047.6
1810.3
1656.4 1570.9
Final
Std 93.131 80.758 71.273 48.040 43.581 57.456 42.802 56.545 53.217
Ave 29.5 57.9 91.7 117.4 137.0 170.9 202.5 223.8 250.1
Deleted
Std 4.720 6.540 8.744 5.190 11.195 13.731 14.183 8.779 17.188
Ave 28.5 55.3 87.4 113.6 132.7 165.2 196.2 217.2 244.2
Rebirth
Std
I
5.523 6.516 8.566 5.481 11.786
12.621
14.070
8.804 16.396
Ave 5999.0 5997.4 5995.7
5996.2
5995.7
5994.3
5993.7
5993.4 5994.1
Final
Std 1.155 1.430 2.003 2.150 2.057 2.791 1.636 2.989 2.079
Ave 62.8 160.4 303.0 467.3 640.3 876.0 1115.0
1319.8 1591.6
Deleted
Std
13.887 12.572 26.546
23.171
20.575
38.335
31.027
37.821 48.761
Ave 60.1 152.8 290.6 449.4 614.4 847.2 1080.2
1282.2 1544.0
II
Rebirth
Std
12.940 13.079 23.524
22.599
20.178
35.925
32.152
31.825 48.475
Ave 5997.3 5992.4 5987.6
5982.1
5974.1
5971.2
5965.2
5962.4 5952.4
Final
Std
1.767 1.713 5.103 2.767 4.654 4.517 5.051 7.058 8.181
4.2 Comparisons When the Self Set Is Unavailable
Another experiment is carried out without the self set being kept, i.e. information about
known self set is unavailable when the detector set needs to be updated. In this case, the
process for generating a new detector in strategy 0 of ASTA-CED [8] can not be
performed here. It is noted that the essential difference between the detector set
updating strategies in [8] and [9] is whether a new detector is tried to be generated when
an old detector is deleted. Therefore, strategy 0 is transformed to the corresponding
strategy in [9] now. For convenience, the detector set updating strategy in [9] is also
denoted as strategy 0 in this subsection. The initial
N
R
is still set to 6000.
Table 3.
Comparison on average
△
C
between the three strategies when the self set is unavailable
during detection. Standard deviations are also listed. Results are got by 10 independent runs. The
initial
N
R
is still set to 6000.
a
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
Ave 3716.6
7524.7 11820.5 16091.9
20500.7
24944.4
29268.6
33805.7 38212.1
0
Std 262.52
330.46 396.11 379.30 776.78 481.05 771.70 785.90 785.43
Ave 2587.8
5095.4 7468.4
9516.9
11432.6
13035.1
14485.4
15906.2 17111.4
I
Std 198.71
189.89 204.72 215.35 355.29 371.95 306.57 314.29 233.93
Ave 1730.4
3522.4 5252.0 6827.9 8534.7 9950.0 11286.9
12893.0 14152.8
II
Std
147.07
85.57 241.13 186.86 213.49 313.52 180.50 329.17 231.70
