Table 4 Statistics on the impact of removing true interactions from the Yeast-D1 Dataset.
Noise | \(Del_{Random}\) | \(Del_{HighDegree}\) | \(Del_{LowDegree}\) | ||||||
|---|---|---|---|---|---|---|---|---|---|
m | \(\vert n \vert _{d=1}\) | \(d_{Avg}\) | m | \(\vert n \vert _{d=1}\) | \(d_{Avg}\) | m | \(\vert n \vert _{d=1}\) | \(d_{Avg}\) | |
0% | 4687 | 28 | 9.4687 | 4687 | 28 | 9.4687 | 4687 | 28 | 9.4687 |
10% | 4249 | 53 | 8.5838 | 4309 | 28 | 8.7050 | 4596 | 32 | 9.2848 |
20% | 3851 | 70 | 7.7797 | 3931 | 28 | 7.9414 | 4506 | 36 | 9.1030 |
30% | 3480 | 89 | 7.0303 | 3553 | 29 | 7.1777 | 4415 | 43 | 8.9191 |
40% | 3191 | 119 | 6.4464 | 3175 | 30 | 6.4141 | 4325 | 45 | 8.7373 |
50% | 2899 | 168 | 5.8565 | 2796 | 35 | 5.6484 | 4234 | 48 | 8.5535 |
