Table. 1 Model performance comparison against seven morphological features and the top three features used for stomata and chloroplast classifications

From: Spatial patterning of chloroplasts and stomata in developing cacao leaves

 

Train set

Test set

 

Accuracy (%)

F1

Precision

recall

Accuracy (%)

F1

Precision

recall

Stomata with full-features

100

1

1

1

81.27

0.81

0.86

0.81

Stomata with selected features

96.71

0.97

0.97

0.97

77.69

0.77

0.82

0.78

Chloroplast with full-features

100

1

1

1

76.96

0.76

0.84

0.77

Chloroplast with selected features

100

1

1

1

76.96

0.76

0.84

0.77

  1. This table compares the performance of the SVM-RBF models for classifying stomata and chloroplast variations using two feature sets: all seven morphological features and the top three features selected by ANOVA. Performance metrics (accuracy, F1-score, precision, and recall) are reported for both the training and test sets.
Back to article page