Table 1.
Summary of PLS-DA results based on all training sets (ER = cross validation error rate in %, LV = number of latent variables, P = evaluation permutation test with e = excellent, g = good, m = moderate, b = bad)
| Model | Training data set | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 4040 | 3030 | 2020 | 1010 | 0505 | |||||||||||
| ER | LV | P | ER | LV | P | ER | LV | P | ER | LV | P | ER | LV | P | |
| 1 | 12.5 | 6 | e | 20.0 | 7 | e | 17.5 | 2 | g | 25.0 | 4 | m | 10.0 | 4 | m |
| 2 | 16.3 | 7 | e | 15.0 | 6 | e | 25.0 | 6 | g | 10.0 | 3 | g | 20.0 | 4 | m |
| 3 | 13.8 | 7 | e | 18.3 | 7 | e | 5.0 | 11 | g | 10.0 | 8 | g | 20.0 | 4 | b |
| 4 | 12.5 | 7 | e | 18.3 | 7 | e | 22.5 | 6 | g | 20.0 | 3 | m | 60.0 | 3 | b |
| 5 | 11.3 | 6 | e | 18.3 | 7 | e | 12.5 | 10 | g | 10.0 | 2 | g | 50.0 | 1 | b |
| 6 | 15.0 | 7 | e | 8.3 | 7 | e | 20.0 | 6 | e | 10.0 | 2 | g | 40.0 | 1 | b |
| 7 | 15.0 | 7 | e | 15.0 | 7 | e | 20.0 | 8 | g | 15.0 | 3 | m | 20.0 | 1 | m |
| 8 | 11.3 | 6 | e | 16.7 | 7 | e | 17.5 | 6 | g | 25.0 | 1 | m | 40.0 | 1 | b |
| 9 | 16.3 | 7 | e | 16.7 | 7 | e | 35.0 | 6 | g | 50.0 | 5 | b | 40.0 | 2 | b |
| 10 | 7.5 | 8 | e | 13.3 | 6 | e | 10.0 | 9 | g | 20.0 | 5 | m | 0.0 | 1 | g |
| Mean | 13.1 | 7.0 | 16.0 | 7.0 | 18.5 | 7.0 | 19.5 | 4.0 | 30.0 | 2.0 | |||||
| SD | 2.7 | 3.4 | 8.3 | 12.3 | 18.9 | ||||||||||