Skip to main content
. 2022 Jun 3;17(6):e0269361. doi: 10.1371/journal.pone.0269361

Table 3. Cross-validation results of Support Vector Machine to predict pollen type, based on 99 iterations of random 80:20 splits of data into training:test sets.

Model parameters precision rate ± s.d.
50 parameters, polar view and equatorial view 96.1% ± 0.3
50 parameters, polar view only 97.2% ± 0.3
7 parameters, polar view and equatorial view 94.7% ± 0.3
7 parameters, polar view only 95.9% ± 0.3