Skip to main content

Table 9 Average number of true-positive and false-positive connections for three algorithms

From: Learning restricted Boolean network model by time-series data

K Noise (%) Algorithm m = 10 m = 20 m = 30 m = 40
TP FP TP FP TP FP TP FP
3 0 Three-rule 6.2 0 8.7 0.6 11.3 1.6 13.3 3.0
New 8.7 3.1 10.5 3.1 11.8 3.3 12.5 3.3
Best-fit 8.1 4.6 10.2 5.4 12.2 6.4 13.3 7.0
5 Three-rule 2.6 2.7 7.3 11.5 10.6 20.7 12.5 30.3
New 7.0 7.5 8.7 6.9 10.1 6.3 10.7 6.3
Best-fit 7.1 11.1 9.2 15.1 10.8 15.7 11.6 15.9
10 Three-rule 1.8 3.6 6.5 17.6 10.5 31.6 12.4 39.8
New 5.5 10.0 6.9 9.5 8.1 9.2 8.4 9.1
Best-fit 6.0 15.2 8.1 19.1 9.2 19.3 9.9 19.0
5 0 Three-rule 6.7 0.1 8.9 0.6 11.0 1.3 12.6 2.3
New 8.3 2.7 9.9 3.0 10.9 3.4 11.4 3.9
Best-fit 8.2 4.6 10.1 5.4 11.8 6.4 12.7 6.9
5 Three-rule 3.0 3.2 7.86 11.8 10.7 20.5 12.8 28.6
New 6.7 7.6 8.4 7.0 9.3 6.7 9.8 6.3
Best-fit 7.1 11.5 9.2 15.4 10.4 15.7 11.1 16.1
10 Three-rule 2.7 2.8 6.9 16.5 10.6 31.6 12.4 39.4
New 5.3 9.9 7.0 9.5 7.5 9.3 8.1 9.1
Best-fit 7.2 11.5 8.2 18.9 9.0 19.3 9.4 19.4