Algorithm #

Size (no. of time points)

Network connectivity ratio (%) - number of non-zero parameters

Miss / total ratio

Incorrect / total ratio

Rank (norm) score

1

4

13

0.94167

0.14167

0.9575

2

4

13

0.94167

0.14167

0.9575

1

5

16

0.63918

0.14145

0.65904

2

5

16

0.63918

0.14145

0.65904

1

6

24

0.48424

0.33622

0.59018

2

6

24

0.48424

0.33622

0.59018

1

7

23

0.34182

0.17576

0.38604

2

7

23

0.34182

0.17576

0.38604

1

8

23

0.23455

0.097917

0.25519

2

8

23

0.23455

0.097917

0.25519

1

9

23

0.14268

0.032794

0.14787

2

9

23

0.14268

0.032794

0.14787

1

10

24

0.1063

0.043982

0.11522

2

10

24

0.10797

0.03455

0.11377

Results of transposive regression (#1) and repressive regression (#2) algorithms for a range of numbers of time points for networks of 10 interacting components. 100 network instances each of seven (7) different datasets are simulated and tested using performance criteria established in section 4.5.
Table 1: Summary statistics of algorithm performances.