Table 1.
Quantitative comparison of the performance evidence for each swarm prediction strategy discussed in this paper.
| Quantitative criterion | Strategy 1 – Instantaneous Spectra | Strategy 2 – Spectral Evolution | ||
|---|---|---|---|---|
| Averaged predictive warning | 22 days | 10.1 days | ||
| Averaged alarm errors (number of colonies which produced false negatives (swarming) or false positive (non-swarming) when using the mean criterion value) | Swarming Colonies | Non-Swarming Colonies | Swarming Colonies | Non-Swarming Colonies |
| 3/18 | 0/14 | 5/18 | 2/14 | |
| Raw alarm false positives (number of colonies with hourly alarm triggers when considered not in a state intending to swarm) | Swarming Colonies | Non-Swarming Colonies | Swarming Colonies | Non-Swarming Colonies |
| 18/18 | 14/14 | 18/18 | 14/14 | |
| Raw alarm false negatives (number of colonies with no hourly alarm triggers when considered in a state of intending to swarm) | Swarming Colonies | Non-Swarming Colonies | Swarming Colonies | Non-Swarming Colonies |
| 2/18 | N/A | 3/18 | N/A | |