Skip to main content

Advertisement

Table 5 A comparison of three classification methods

From: A high-throughput stereo-imaging system for quantifying rape leaf traits during the seedling stage

Methods of classification Predicted group accuracy Leave-one-out cross-validated (LOO-CV)
Mosaic Semi-mosaic Round Total (n = 399)
Stepwise discriminant analysis 98.3% 90.3% 100.0% 92.7% 94.4%
Support vector machine (SVM) 100% 92.4% 92.6% 94.7% 95.6%
Random forests classifier 99.2% 87.5% 89.6% 91.7% 94.8%
  1. Fixed half selected for training (402), and the other half (399) for testing