Table 3 Classification accuracy of machine learning algorithms using landmark and semi-landmark data
From: An image database of Drosophila melanogaster wings for phenomic and biometric analysis
Algorithm | Sex (± Standard error) | Genotype (± Standard error) |
|---|---|---|
LDA | 98.2 % (±1.6) | 86.1 % (±1.5) |
QDA | 81.5 % (±6.4) | 68.7 % (±2.2) |
FDA | 98.2 % (±1.6) | 86.0 % (±1.5) |
MDA | 98.1 % (±1.6) | 84.8 % (±1.6) |
Bagging | 93.3 % (±2.9) | 57.6 % (±2.9) |
Random forest | 94.6 % (±2.7) 100 trees | 74.9 % (±2.1) 1,000 trees |
SVM | 96.8 % (±2.1) sigmoid | 83.8 % (±1.6) radial |
Neural network (size 10) | 98.3 % (±1.6) | 81.2 % (±2.2) |
KNN | 98.3 % (±1.5) k = 4 | 59.3 % (±2.1) k = 32 |