Skip to main content
. 2021 Oct 20;8:755007. doi: 10.3389/fnut.2021.755007

Table 2.

Performance of the classification models based on multiple-modality features.

Types of Data No. of Feature Variables Mean Accuracy (%) of Cross-Validation Accuracy (%) of Prediction
LSVM LDA RF LSVM LDA RF
MS + Texture 30 84.25 88.06 75.75 86.36 83.64 80.00
CIELAB + Texture 27 74.83 74.13 71.92 78.18 84.55 61.62
MS +Texture + CIELAB 33 85.37 89.42 80.69 90.00 90.91 80.00