Prediction of plant lncRNA by ensemble machine learning classifiers

BMC Genomics

Table 3 Evaluation measures of random forest (RF) and gradient boosting (GB) ensemble models

ML model type	Ensemble type	AUC	MCC	Accuracy	Sensitivity	Specificity
RF
	Vote	0.834	0.725	0.944	0.594	0.995
	Arithmetic mean	0.963	0.661	0.941	0.562	0.996
	Geometric mean	0.963	0.706	0.941	0.555	0.997
	Logistic regression	0.835	0.765	0.952	0.665	0.994
GB
	Vote	0.887	0.797	0.958	0.702	0.995
	Arithmetic mean	0.945	0.786	0.956	0.681	0.996
	Geometric mean	0.940	0.750	0.949	0.601	0.999
	Logistic regression	0.883	0.822	0.963	0.745	0.994

Statistics for vote, arithmetic mean, and geometric mean models were calculated using outputs of models compared to true labels. Logistic regression evaluation statistics were calculated using the scores found by 10-fold cross validation of O. sativa training data and validated lncRNA sequences

ISSN: 1471-2164