Table 1 The educated machine algorithm's performance (run parameters: 7 iterations, p = 0.15, voting between 5 data cubes).
From: Hyperspectral imaging for chemicals identification: a human-inspired machine learning approach
Feature | Material | Voting applied? | |||
|---|---|---|---|---|---|
Sugar | Silicone | Polystyrene | Total | ||
Probability for detection, % | 89 ± 0 | 98 ± 2 | 100 ± 0 | 96 ± 1 | Yes |
NWIP | 13 ± 11 | 5 ± 3 | 20 ± 20 | 39 ± 21 | |
Probability for detection, % | 95 ± 5 | 98 ± 2 | 100 ± 0 | 98 ± 2 | No |
NWIP | 50 ± 78 | 8 ± 5 | 52 ± 94 | 110 ± 115 | |