Skip to main content

Table 4 Results for machine learning algorithm model accuracies developed using the complete set of iron deficiency chlorosis data on a diverse set of soybean accessions

From: A real-time phenotyping framework using machine learning for plant stress severity rating in soybean

Algorithm

Accuracy

MPCAa

Cross validated MPCA

Interpretability

Cost metric

CT

99.7

91.7

78.4

Low

0.0027

Hierarchyb

99.2

90.7

79.2

High

0.0082

Hierarchyc

98.3

84.0

79.0

High

0.0201

QDA

98.5

83.2

77.9

Medium

0.0201

NB

98.4

79.0

78.5

Medium

0.0284

KNN

99.5

75.8

84.3

Low

0.0073

RF

99.1

75.0

81.1

Low

0.0092

GMMB

99.4

74.2

82.7

Low

0.0064

LDA

98.5

71.7

76.9

High

0.0156

SVM

97.3

45.8

45.3

Low

0.0458

  1. aMean per class accuracy
  2. bSVM: using SVM for both classifiers
  3. cLDA and SVM