Prediction of plant-derived xenomiRs from plant miRNA sequences using random forest and one-dimensional convolutional neural network models

Table 3 Accuracy comparison between RF model and 1D-CNN model on test set and 5-fold cross-validation set

Model		SN	SP	ACC	AUC	MCC
Independent test set	RF	0.920	0.560	0.740	0.866	0.547
Independent test set	1D-CNN	0.880	0.680	0.780	0.817	0.574
5-fold CV	RF	0.901	0.666	0.736	0.840	0.538
5-fold CV	1D-CNN	0.859	0.712	0.766	0.794	0.543

ACC, SN, SP, AUC and MCC indicate accuracy, sensitivity, specificity, area under the ROC
curve and Mathews correlation coefficient, respectively

ISSN: 1471-2164